Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
Proposal for a collaborative learning system

By Stick in Internet
Sun Sep 28, 2003 at 06:45:21 AM EST
Tags: Round Table (all tags)
Round Table

As a self taught programmer I have always been interested in finding better and faster ways of educating myself on any subject. In my experience I have found the biggest blockade to learning is the quality of information. The problems I most commonly encounter are:

  • Assumes you already have studied a certain technique/skill/area but fails to mention this. This can make many people give up as the subject seems difficult when actually they are just missing a key piece of information essential to understanding it.
  • Wastes time explaining a different subject to the one it is meant to be explaining. The worst example of this was a tutorial on encryption I once read where 90% of it explained how to code in C (badly), and then only spent a few lines at the end explaining encryption. For people who already know C this tutorial was 90% redundant.
  • There is no structure across different documents which often leaves gapping holes in a students knowledge. These holes often lead to the first problem described occurring.
While these problems won't stop a determined individual from learning it does waste time and will almost certainly leave some holes in the individuals knowledge. I have come across these problems many times, and while my ability to work around them has improved greatly I still feel that I and many others are wasting too much of our time battling with them.


For the last two years I have thought about ways of solving these problems. The ideal solution would be for a single person to write a cohesive set of documents starting from the absolute beginner stepping right through to expert without leaving any gaps in the information. However, I highly doubt a single person skilled enough for this task exists and if they did I also doubt that they would have the time to do such a large task. For example, imagine how long it would take for one person to write up just one segment of the IT field (e.g. Computer Science) as it stands today, never mind keep up with new developments.

However, if this set of cohesive documents were written by a variety of people this would solve the problems associated with a single person (as well as introducing some of its own). Kuro5hin is an example of collaborative media that we're all familiar with. I believe a system similar to Scoop (the software Kuro5hin runs on) is the right direction. A web based system that allows users to create content and is moderated by peer review. Scoop is not an ideal an ideal system for solving the problems outlined earlier. I believe that a custom system with the following features is required.

  • Topics broken into modular lessons - Each topic of information should be covered in its own lesson. For example, one may be tempted to place the topic of C1 programming and the topic of linked lists in the one lesson due to their close relation however it would be more appropriate to give them separate lessons. Not everyone who wants to learn C wants to learn how to create a linked list2 and vice versa. The real benefit of this is seen when we look at the overall structure of the site in the next point.
  • Tree like structure organisation of information - Lessons should start with with core concepts then branch out into different areas. This allows beginners to the subject start from the bottom up. For example, a website teaching German to English people would start off by explaining how language is actually constructed (sentences, paragraphs, vowels, nouns etc.). The next lesson in the tree would then cover how language construction differs in German. Then it could branch of into topics such as writing in German or Speaking in German. The user can choose a path of learning that suits their current interests. Users more experienced in the subject can easily enter the tree at a level they are comfortable with. Because the lessons are modular they won't waste time reading redundant (to them at least) information. The tree structure allows us to present all the information to become proficient in field while still giving them the freedom to choose which direction is takes.
  • Defining assumptions - Each lesson should clearly define what knowledge is required to fully understand the information and ideally link back to lessons that contain that knowledge. If a user finds that what they are reading is above their level they can easily traverse back down the tree until they reach a comfortable level.
  • Peer review - When someone submits a new lesson is should be peer reviewed much like the edit queue on Kuro5hin. However, unlike Kuro5hin an unspecified amount of time should be given to review the article in order allow users to do their own research for the article. Users should be encouraged to contribute as well as fix mistakes. Rather than a voting system users would have a flag they could turn on or off when they felt the lesson was complete. Once complete an editor would check the lesson and add it to the tree. Having editors involved ensure that the information meets the websites quality standards as well as prevent abuse of the flag system.
  • Lesson specific comments - Users should be able to attach comments relating to the lesson. The online PHP3 documentation already has this feature and it has proved extremely useful. User comments often provide practical advice relating to real world issues and problems.
  • Regular lesson reviews - In most fields there are many developments and changes over time. In the IT world for example languages has such as C have changed quite a bit since they were first developed. It is important that lessons undergo regular reviews to ensure that they are still relevant in the future.
  • Strict editorial guidelines - Because many lessons will be written by different authors the possibility exists for each lesson to have it's own writing style and format. A familiar style and format is easier for the user to learn from so it is important that writing styles and lesson formats are clearly defined. What these are will depend on the subject.
  • Funding - An issue often neglected with many websites is the financial side of things. Large websites cost money to run in terms of bandwidth, hardware, and server administrators. Contrary to popular belief there is actually quite a lot of money in online advertising. The forum webhostingtalk.com reportedly has an income of over $200,000 per month. Provision should be made for the website be financially viable in the long term.
The end goal is to have a series of well written lessons that when put together give complete instruction in the entire field and is updated and improved regularly to ensure it remains relevant for years to come.

Ideally the best way to implement something like this would be as an Open Source project like Scoop. This would mean that people from non-tech related fields could implement their own versions. For example, there could be systems to teach math, languages, the various sciences, and much more.

--------------------------------------
1 The C programming language.
2 Linked lists are structures used to hold data in programming languages. They are mainly used in C programming.
3 The PHP scripting language

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Related Links
o Scoop
o Kuro5hin
o webhosting talk.com
o The C programming language
o Linked lists
o The PHP scripting language
o Also by Stick


Display: Sort:
Proposal for a collaborative learning system | 136 comments (122 topical, 14 editorial, 0 hidden)
This is a good idea. (4.66 / 9) (#1)
by razumiking on Sat Sep 27, 2003 at 02:56:52 PM EST

There should be a cohesive set of documents to explain the various aspects of programming, software development, systems administration, and so forth. It would surprise me if no attempt has been made.

I think, though, that you would need the backing of a well established organization to make this sort of thing happen. I think the Free Software Foundation would be a ideal here. In Stallman's original GNU proposal he alludes to a massive documentation effort, so this may well be in line with the mission of FSF. Also, since these documents could be molded into textbooks and published by the FSF, we could expect this to be a source of income and notoriety for the FSF. Certainly with the way Sun and Microsoft have insinuated themselves into the training of High School and College level education, it would be a good idea for the FSF to keep pace.


Donations (4.25 / 4) (#61)
by ZorbaTHut on Sun Sep 28, 2003 at 07:43:58 PM EST

Be careful of taking money from corporations with goals. I imagine everyone on this site knows what happens when Microsoft funds a usability study, or the tobacco industry funds a cancer study, or (political_group_of_your_choice) funds a (hotly_contested_subject) study.

While many people on this site might agree with the FSF, letting them influence the site would cause an instant loss of credibility among skeptics. I personally think free software is a great thing, and loathe the way the FSF is going about it - so if I start seeing lots of "Free software is the best! Here's how to do it!" papers springing up, without a single "how to program Windows applications" tutorial, I'm going to think that *somebody's* been paid off.

[ Parent ]

knowledge holes (4.76 / 13) (#3)
by adiffer on Sat Sep 27, 2003 at 03:19:24 PM EST

I spent a few years learning how to learn in order to get through grad school without failing out.  The other guys simply learned faster and I did so I tended to land on the bottom of the curve.  To survive I had to figure out what they did that I did not.  I managed to find it and restructure my learning effort a bit, but not enough to make a really big difference to my grades.  I did survive though.

Along the way, I learned something very useful that wasn't on my goals list.  I learned that those knowledge holes actually prove useful to the evolution of knowledge as a whole.  Think of them like mistakes in your genetic code and you'll get an idea of what I learned.  Every one of us fabricates ways to know how to do a variety of things.  Because of the mistakes and gaps in our knowledge, each approach can be treated as though it had a genetic source.  Some of the end results thrive and others don't.  Occasionally a mutation occurs, though, and what used to be considered knowledge gets dropped or changed.

I'm sure most fields require a great deal of the knowledge you have to be good and self-consistent.  Little animals don't get along well, after all, if they can't even make proteins.  However, don't be too harsh on the mistakes and gaps left by the chaotic learning process.  They are the very things that make it possible for the state of knowledge to advance.  Make sure they are small enough to ensure you can survive and grow, but don't strive TOO much for conformity and order.
--Alt.Space Conference (SFC 12)

i think people should heed your words. [nt] (5.00 / 2) (#12)
by Prophet themusicgod1 on Sat Sep 27, 2003 at 06:33:45 PM EST

i could go on about how there really is little justification in all our knowledge claims, so building this giant learning tree would be in vain...

but i won't, for now..
"I suspect the best way to deal with procrastination is to put off the procrastination itself until later. I've been meaning to try this, but haven't gotten around to it yet."swr
[ Parent ]
justification?! (4.33 / 3) (#34)
by adiffer on Sun Sep 28, 2003 at 01:45:30 AM EST

Knowledge needs justification?

8)

It's just the yin and yang thing again.  Those who bring order to the system do some good and then discover in the midst of their greatest work they have also created chaos.
--Alt.Space Conference (SFC 12)
[ Parent ]

Only if its misaligned with the above knowledge.nt (4.33 / 3) (#77)
by Russell Dovey on Mon Sep 29, 2003 at 05:31:41 AM EST


"Blessed are the cracked, for they let in the light." - Spike Milligan
[ Parent ]

or (4.33 / 3) (#93)
by Prophet themusicgod1 on Mon Sep 29, 2003 at 12:38:50 PM EST

the chaos was there all along, we just have to acknowledge it.
"I suspect the best way to deal with procrastination is to put off the procrastination itself until later. I've been meaning to try this, but haven't gotten around to it yet."swr
[ Parent ]
tree of knowledge (4.78 / 14) (#4)
by speek on Sat Sep 27, 2003 at 04:13:35 PM EST

I like the idea of the tree. Unfortunately, I think you'll find many lessons, of not most, have multiple "parents", or prerequisites. Your tree then turns into a network of nodes organized in layers, each layer containing nodes that point to one or more nodes in the previous layer. And, since there's no reason to think that any given node's parents will exist only in the previous layer, we have to again expand the complexity and allow nodes to point to parents in any previous layer. At that point, we probably have to give up on the notion that we can categorize our nodes into "layers", and instead allow a free-flowing graph of nodes that hopefully have no loops, orphans, or conflicting dependencies. Fortunately, I don't see that as a problem.

Even so, it seems like a useful idea to implement somehow. I especially like the option for commentary on any given node.

--
al queda is kicking themsleves for not knowing about the levees

web of knowledge (4.86 / 15) (#6)
by celeriac on Sat Sep 27, 2003 at 04:33:09 PM EST

Lots of people have tried to organize knowledge into a kind of tree structure (just look at library catalogs for instance) but every time someone nails down a tree, a new field pops up that is a combination of things from very different branches of the tree. You wind up with a third the books related to a particular subject living on one floor of the library, another third on another floor, and the remaining third scattered throughout.

Knowledge tends to be more of an amorphous web, where each concept builds on a small group of antecedents. But once you get far enough out it becomes impossible to tell exactly what concepts are dependent on others--often you find that you can combine concepts A, and B to synthesize C, but sometimes it's just as easy to synthesize A from C and B. Then you wind up with some people who learned C and B in school, and other people who took a different route and picked up A and B--they're building knowledge webs in different but equally valid ways, and they need different kinds of lessons in order to progress.

The idea of a collaboratively-formed knowledge base is compelling. I've found it's much easier to pick up mathematical concepts from a site like Mathworld.com or PlanetMath.org, which tend not to have much of an overall structure, but manage to link together all the requisite concepts so that they represent the web of knowledge around any particular concept. Wtih a tabbed browser I can start at the page I want to learn, click the first link I'm not familiar with, keeep drilling down util I find concepts I understand, and eventually come to an understanding of what I wanted.

PlanetMath.org has a very interesting base of software running it, Noosphere, which is possibly the first collaborative-authorship scheme I've seen that actually seems to work. It seems like Noosphere would be able to cover many of the requirements you bring up as far as peer review, editorial standards, linking to requisite concepts, fine granularity of knowledge, and so on. I'm planning on starting a personal website running Noosphere soon, to see how it works for organizing more general content.

Thoughts on meta data (4.22 / 9) (#8)
by Tatarigami on Sat Sep 27, 2003 at 05:18:39 PM EST

I think that ideally, every module should be accompanied by the following info:
  • contents - what the module should teach you, preferably described in generic terms so that complete beginners can see where it fits into their overall study goals
  • more contents - a specific (but short) list of what's in the module (eg. 'introductory grammar' for a language, 'file access methods' for programming, 'deep frying techniques' for cooking). There's no reason a very basic module couldn't also be a useful reference for someone experienced in the subject.
  • pre-requisites - what you should already know before beginning this lesson
  • side requirements - what you should be studying at the same time as this lesson
  • estimated time for completion


Side Requirements (4.80 / 5) (#55)
by MyrddinE on Sun Sep 28, 2003 at 03:13:19 PM EST

No such thing as 'Side Requirements'. There should NEVER be another lesson you should be taking at the same time as a certain lesson... if there was, then it should be in the same lesson.

A better description would be 'loosely coupled' topics. For example, when discussing arrays and linked lists, I do not need to understand binary trees. And binary trees have little to do with arrays (though they are related to linked lists). I could easily write a lesson on arrays, one on linked lists, and another on binary trees that do NOT depend on each other for understanding.

If they were written so they could be learned independantly, then it would be possible to read them in any order. They could have notations in them that they are 'loosely coupled' with the other data structure topics, but the other topics are not REQUIRED.

A higher topic may require that I understand arrays and linked lists. Another upper level topic might require I understand binary trees. So in the end I will need to learn them all for proper understanding. But while learning about linked lists, the other data structures are just 'loosely coupled' to each other, in that they are all related and reading them together may aid my understanding.

This is different than your idea because they are not 'requirements'. Merely suggestions. If another topic is a 'requirement' then you have to know it FIRST. That's the whole definition of a requirement.

[ Parent ]

For this to work (4.36 / 11) (#11)
by godix on Sat Sep 27, 2003 at 06:17:59 PM EST

You would need one hell of an editor. Before a lesson could be written someone has to sit down and figure out how the subject breaks down and what the individual lessons should be on. The editor, preferable the same person who broke out the lessons, would then have to proofread the lessons submitted to make sure they actually covered the topic assigned and didn't make assumptions in knowledge. Sites like K5 can survive if the man in charge has other things occupying his time and attention, your proposed site would be absolutely devestated by the same absence. For that reason I suggest the editor in chief (or whatever you want to call him) be a paid position and have the same expectations of attendence and job performance as a regular job would have. Volunteer hobbyist have a habit of eventually getting bored and moving on.

The other vitally important thing is censorship. Your site would have to make quite clear that the lessons are NOT a chat forum and any comments even slightly offtopic will be moved or deleted. Consider K5, if this article makes it out of the queue I expect at least 50% of the comments will not be related, or at best very very loosely related, to the topic. If this was allowed in a lesson readers would get bored and leave or spend they could use to learn sifting through 'how's it going' type crap instead. You could of course have nonlesson forums for chats, but whoever ran the site would have to be a censoring dictator in the lessons.

I don't understand spending all that money for a fancy shot ... when pregnancy ain't nothing that a good coathanger or a pair of steel toed boots can't fix<

I agree totally (4.50 / 4) (#16)
by Stick on Sat Sep 27, 2003 at 07:15:36 PM EST

A project of this sort would be quite an undertaking and require a lot of discipline. That's one reason why I mentioned funding as one aspect. We can see the problems K5 has had as result of not enough money being available. Just keeping the site online proves can prove to be a major undertaking in itself.

I agree with the censorship aspect as well. I really should have mentioned in the article actually. Having useless information strewn around goes completely against the goals of the project.


---
Stick, thine posts bring light to mine eyes, tingles to my loins. Yea, each moment I sit, my monitor before me, waiting, yearning, needing your prose to make the moment complete. - Joh3n
[ Parent ]

Fortunately (4.60 / 5) (#29)
by rusty on Sat Sep 27, 2003 at 10:58:44 PM EST

A project like this is blessed with a pretty clear goal and program, so that kind of strict editing would be expected and probably desired by most people. Cries of "censorship!" ought to be fairly dismissable.

One question that occurs to me is whether you've thought of this in the realm of strictly technical knowlege, or as a general platform for learning. Because when you move into the fuzzier subjects like history, for example, then you get into issues like whether some article was deleted because it was a bad lesson, or because it didn't agree with the editor's notions of what the "truth" is. Something to think about.

I also think planning for a paid administrator would be a wise decision. Even if they are mainly serving a managerial role, wrangling volunteer editors and moderators and riding herd on the hosting company, it's a lot of work and very little of it enjoyable. People will volunteer to do the glamorous work of being the Guy In Charge and Resident Expert for some topic, but there's gotta be someone doing the unglamorous and usually unpleasant work of coordinating them and making sure everyone's actually doing what they're supposed to be.

____
Not the real rusty
[ Parent ]

Wiki-like aspect (4.75 / 4) (#45)
by gyan on Sun Sep 28, 2003 at 09:37:38 AM EST

Because when you move into the fuzzier subjects like history, for example, then you get into issues like whether some article was deleted because it was a bad lesson, or because it didn't agree with the editor's notions of what the "truth" is. Something to think about.

 You could have a "lesson discussion area" where all authors cite their sources, argue about it seriously and then hold a vote, if need be.

********************************

[ Parent ]

Epistemology and ontology (4.33 / 3) (#53)
by dollyknot on Sun Sep 28, 2003 at 02:30:51 PM EST

Epistemology: The world inside your head

Ontology: The world your head is inside of.

Peter.


They call it an elephant's trunk, whereas it is in fact an elephant's nose - a nose by any other name would smell as sweetly.
[ Parent ]

I disagree; (2.00 / 1) (#115)
by snowlion on Tue Sep 30, 2003 at 11:11:34 PM EST

Look at the wikipedia entry on Capitalism.

I think that's remarkably fair- don't you?

Do you think it'd be fairer if there were an "official moderator" or "official editor?"
--
Map Your Thoughts
[ Parent ]

I'm not sure I follow (3.00 / 1) (#117)
by rusty on Wed Oct 01, 2003 at 07:36:25 AM EST

Are you responding to the first part of my comment, about strict editing, or the last part about having a full-time admin? Or, I guess, the middle part?

____
Not the real rusty
[ Parent ]
Obligatory link... (3.85 / 7) (#18)
by Pseudonym on Sat Sep 27, 2003 at 07:32:13 PM EST

SCORM



sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
Why mod down ? (4.33 / 3) (#23)
by IriseLenoir on Sat Sep 27, 2003 at 09:37:02 PM EST

Although the site's design seems childish, the content is serious and definitively on topic. Sure, some content only works on I.E., it's by the U.S. Department of Defense who's goal is "creating a vigorous market for these products" and "a learning economy", the other projects mentioned seem much more interesting, but still... I only scanned the start of the guide, would the poster of the link care the resume the project for us? I'll mod your post back up if you ;-)
"liberty is the mother of order, not its daughter" - Pierre-Joseph Proudhon
[ Parent ]
It lacks content. (4.50 / 4) (#38)
by razumiking on Sun Sep 28, 2003 at 03:52:16 AM EST

An explanation of the link would be good. The page linked to does not give a good enough description of the project for the link to be interesting without a bit of background added. It doesn't make sense to link to a site that someone will have dig around through without adding some sort of explanation.

[ Parent ]
Well... (4.25 / 4) (#67)
by Pseudonym on Sun Sep 28, 2003 at 09:34:26 PM EST

SCORM is kinda hard to explain. The short version is that it's supposed to be an interoperable standard for sharing reusable course modules.

The real story is that it's actually nowhere as "reusable" as the designers claim. The big problem is that in SCORM, you share HTML fragments, and we all know that HTML has big problems with separating content from presentation. This means that a module written by one person really can't be imported into a course prepared by someone else if the course creator expects the module to be consistent.

If that issue can be overcome (e.g. using XML for the repository and letting everyone use their own stylesheets/XSLT for presentation), I think that something like SCORM is the Right Thing(tm).



sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
[ Parent ]
SCORM in a nutshell (4.33 / 3) (#74)
by kubalaa on Mon Sep 29, 2003 at 01:05:36 AM EST

SCORM has nothing to do with structuring knowledge or learning in a uniform way. It's all about the concrete details of taking web courses and being able to glue them awkwardly together in different learning management systems (LMSs).

SCORM has two main parts:

  1. XML schemas for defining the structure and content of course packages and metadata, so that an LMS can describe courses in its catalog and give users a very high-level table of contents.
  2. A "runtime" API which is a set of javascript functions used by SCORM courses to send student-specific data like test scores to the LMS.
One drawback (or advantage) to SCORM is that it is not very fine-grained. Most of the interesting things that go on inside web courses, like navigating from page to page, or taking tests, are not specified by SCORM and have to be done by the course author.


[ Parent ]
Some of what you need -repost as topical (4.71 / 7) (#19)
by minerboy on Sat Sep 27, 2003 at 08:33:30 PM EST

This is an interesting Idea, and one I have thought about also - though not just for teaching programming - here are some things I think such a system needs

  • Users of such a Scoop site would not submit Articles, they would submit learning objects - the site must be set up to allow a user to submit these in some standard form without learning exhaustive sets of standards
  • The Scoop system needs to be arranged using metadata that will appropriately link the various learning objects that are submitted by users that would have some interoperabillity usingIMS standards for example
  • There needs to be some kind of self-testing system to give the user feedback about how well thay have learned a topic.
  • There has to be a distinction somehow between teacher and learner. To me this would be one of the most difficult parts, because someone could be a teacher for introductory things ( like Using HTML) and be a learner for more difficult areas (like relational database programming, or Quantum mechanics) at the same time there is a very important distinction between the comments/ratings from a students, and from an instructor
  • The distinction also needs to be made for interactions between users, so one knows whether they are student - teacher or student - student or teacher -teacher
  • Actually, I don't like the teacher - student distinction at all, maybe there should be expertise levels or something for users - seems more democratic somehow
  • One last item, there would need to be a start up period when only content creators would be using the site - until a core set of material was together, and made ready to use


Wow ! (4.50 / 4) (#21)
by IriseLenoir on Sat Sep 27, 2003 at 09:23:24 PM EST

I've thought about this a bit, and although I was sure I wasn't the first to have such an idea, I didn't think there was this amount of research made on this.

My idea was just this - reusable learning objects that could be automatically arranged to make a personalised lesson based on some information provided by the user - what does he want to learn, where does he start from, what form of presentation does he best learns from, what are his primary interests and areas of knowledge (so examples and metaphors from this domain could be used, thus making it more easy for him to connect the knowledge with things he is familiar with), etc.

That information would be stored and reused for other lessons of course, and a system for semi-automatically updating it could be used, based on the lessons he learned from and feedback from the user on how much information he retained from it (and a system for testing this, as you said). User could also add knowledge he learned elsewhere in the base, eventually building a comprehensive database of the user's knowledge. This could also be used to make quick refresher lessons of things he's supposed to know to ensure it isn't forgotten from lack of use. And even, optionally, to permit others to search other people's metadata, looking for help on a particular topic, looking to employ people with specific knowledge, or even looking to communicate with people with similar interests (whether to debate topics and share new information, or make friends.)

Dependencies he already learned would thus not appear in a lesson, but would still remain easily accessible from it for refresher, or if the user's metadata is somehow wrong. This could be done, for example, through links that would expand an in-page paragraph briefly explaining a concept the first time it is referenced and containing other links that would allow to remove the concept from the knowledge base, open (create) another lesson on it, list the concept's dependencies to make sure it isn't because of another hole that you don't understand it in this specific context, etc.

I wrote a small essay on this some time ago, sad i can't seem to find it. I'll post it here later if I do...

"liberty is the mother of order, not its daughter" - Pierre-Joseph Proudhon
[ Parent ]

You really don't know what you're doing (none / 0) (#129)
by Wiggy on Sun Oct 05, 2003 at 10:37:21 AM EST

Quite simply, the study and research of this particular pedagogy is massive.

Here are some reasons why the assumptions you've made fall down in the practical sense:

  • How do you define a user's skills and competencies? Capture a description of everything they've ever done? How does that reflect tacit knowledge which is not easily recorded?
  • How do you map that against what they need to do next? A skills matrix of some sort? Great, you've just killed an entire industry. What I describe as a Sys Admin is not going to be the same as what you describe as a Sys Admin. As a result, the skills each definition requires will change. You've pre-empted and shoeboxed a person into something they're not.
  • How does any of this fit into any CPD or Accreditation model that means the learning actually means something? How do you manage assessment and tracking? Tests? Evidence of Applied Learning? Who monitors what and how? Is learning not learning until you've shown an understanding of it? Is that not why we have exams? Can I not apply the learning somehow instead of being tested? How do I show I know about military strategy after studying a history course in Roman warfare? Test or application? Who sets the tests and makes the decisions in the first place? If I don't do an assessment of some sort, was it not just "reading" instead of "learning"?
  • What format should Reuseable Learning Objects (RLOs) be submitted in? Let's suppose I have data for population broken down by age for Nigeria in 1996. Is it more useful to have the raw data, a graph or a pie chart? Answer: depends on how the RLO will be used. A history course will find a graph fine, a statistics course or economics course will require the raw data. What about a course on how to draw pie charts?
Trust me, your idea isn't novel but has been considered interesting enough that over a dozen European universities are looking at it, and it will happen. But you're about 5 years behind the rest of us. I'm just showing you the very smallest tip of the iceberg here.

RLOs and a working system based on the pedagogy you describe is desirable and being constructed, slowly, elsewhere. I suggest that for at least 12 months you hold back and read a smany research papers as you can.

Searching for "KOD" and "FD Learning" should show you something that's quite interesting as well and takes things in a different direction.

[ Parent ]

What you want is a Wiki. (4.76 / 13) (#22)
by Russell Dovey on Sat Sep 27, 2003 at 09:23:55 PM EST

Wikipedia as a base might be better for this sort of thing than Scoop. The real innovation would be the lesson tree/web.

"Blessed are the cracked, for they let in the light." - Spike Milligan

Wikibooks (4.55 / 9) (#35)
by brion on Sun Sep 28, 2003 at 02:37:39 AM EST

More specifically, check out Wikipedia's sister project Wikibooks.

Chu vi parolas Vikipedion?
[ Parent ]
+1, SP. (2.37 / 8) (#25)
by Akshay on Sat Sep 27, 2003 at 10:08:30 PM EST

Want more comments on this.

Copyright issues (4.66 / 9) (#26)
by Tatarigami on Sat Sep 27, 2003 at 10:15:05 PM EST

Just a side note, but if such a thing were to be done by the K5 community, I'd like to see the content released under something like the Creative Commons Attribution-NonCommercial-ShareAlike licence.

On Kuro5hin, copyright isn't often relevant because -- let's face it -- very little of what we write here has any potential monetary value. On the other hand I think the idea of copyright needs to be brought up, because this article discusses making a resource freely available that people currently pay thousands of dollars for. I personally would not be happy for some fly-by-night training college run out of a garage to take a tutorial I wrote and distributed to show people how clever I am out of community spirit and try to charge money for it.

Why would that be a problem? (4.60 / 5) (#30)
by razumiking on Sat Sep 27, 2003 at 11:08:09 PM EST

I don't see why you would not allow someone to charge money for such a thing. If a community college charges students for a coursebook containing material from a project like the one described, that's actually a good thing because the material is being put to good use. Ideally, the project could put together textbooks of the material and publish them to pay for expenses and so forth.

I would much rather see something like this put under some license that allows for free redistribution for any purpose as long as it isn't blatant exploitation -- GNU Free Documentation license or maybe something less restrictive.

In any case, it is important for a project like this to actually be used. More users will likely lead to more contributors. Also, if professors use these texts in their courses, they will be more likely to contribute themselves, which would obviously be a very good thing. But if, for example, they cannot charge a fee to recover printing fees from the local copy shop, they will be more likely to just use some regular textbook.

[ Parent ]

I should be more specific (4.33 / 3) (#36)
by Tatarigami on Sun Sep 28, 2003 at 03:08:13 AM EST

This sounds like a great idea, and I would be proud to contribute, in the unlikely event that anyone needs to learn my skills. However, in donating my time and knowledge, I'd have an expectation that users would have free access to what I produce.

If anyone intended to profit by my efforts by charging more than their costs for material I'd given away, especially if it was without consulting me first, I'd be less than pleased.

[ Parent ]

I guess I just don't understand that mentality. (4.60 / 5) (#37)
by razumiking on Sun Sep 28, 2003 at 03:42:38 AM EST

If I were to be involved in an undertaking like this one, I would be more interested in the dissemination of the material than in who profits from it. I think there is a certain level of pragmatism that needs to be applied. It's like the plant that spreads its seeds through the defecation of birds. Certainly the bird gets a great deal in the arrangement and the plant plays a rather unglamorous role, but ultimately the seeds are spread which is what the plant wants.

It's really academic on some level though. When things are freely available online, they are usually much cheaper in regular markets. For example, a guy at Cornell named Allen Hatcher wrote a book on Algebraic Topology and makes it freely available online. He also has it published with the Cambridge Press. The book sells for $30 USD (cf. $70-$110 for books on the same topic from Springer and others). As long as the texts are available online and no one is allowed to add their own non-free parts when selling them, the texts will remain reasonably priced in paper form.

If you want a project like this one to be used by people other than dilletante programming dabblers, you will have to make allowances for others to publish and reproduce them (even for profit) or actively print and publish the works yourself.

[ Parent ]

They Would (4.50 / 4) (#42)
by freestylefiend on Sun Sep 28, 2003 at 07:20:18 AM EST

in donating my time and knowledge, I'd have an expectation that users would have free access to what I produce.

The Creative Commons Attribution-ShareAlike license would allow free access to the original as long as the author (or anybody else with a copy) was willing to distribute it for free. Free access to derivative works would usually be available, unless nobody who had access to the modified version was willing to distibute it for free. The share alike part means that those who make commercial use of the material would not have a monopoly over it after they begin distribution.

[ Parent ]

If content is free for these with IQ <140 (2.71 / 7) (#31)
by United Fools on Sun Sep 28, 2003 at 12:16:19 AM EST

We are all for it.
We are united, we are fools, and we are America!
And I thought.... (5.00 / 2) (#33)
by IriseLenoir on Sun Sep 28, 2003 at 01:18:00 AM EST

...your organisation was meant to fight intelligence-based discrimination!
"liberty is the mother of order, not its daughter" - Pierre-Joseph Proudhon
[ Parent ]
This is an interesting idea (4.44 / 9) (#32)
by Michael Moore on Sun Sep 28, 2003 at 12:56:30 AM EST

I think one of the main problems with this sort of thing would be how much knowledge the author of any node will assume the reader has. Something like 'linked lists' could really be anywhere from the third to the fiftieth tutorial a person could study, depending on how the author designed it. I suppose, however, that you could allow some sort of difficulty scale, possibly both set by the author of the piece as well as everyone who views it. Or perhaps force people to specify in the tutorial itself specifically which other tutorials should be required reading (with links or whatever) before attempting this one.

Also, there's the tendency of programmers to have screwy little quirks in their coding style. Even if you enforce strict guidelines with indenting, formatting, etc., there's the problem of how different programmers tackle different tasks from an actual functional standpoint. This can really vary a lot in its severity as well, from programmers who prefer to deal with text-type data as something structured like an array (probably coming from a C-like language) versus programmers who prefer to deal with it as actual strings (probably coming from Perl or another language with powerful string tools), to entirely different programming paradigms, like whether or not to use C++ objects as opposed to simple functions. A person reading a set of tutorials in which all these variables are jumping wildly around the place could definitely have some difficulty.

And, at the same time, you'll be hard-pressed to get the programmers to stop doing what they're doing, assuming you even can. "But C++ is an OO language, why shouldn't I use classes", "C++ doesn't require me to use classes, why should I have to?", etc. And like I said, a lot of people probably couldn't drop the way they code even if they tried, it's just how they've learnt to think in programming terms. But then, *should* you be trying, with some base-level guidelines for things like not using OO when not necessary? It might even fragment a tree into "Bill's OO C++ tutorial set" and "Bob's old-school C++ tutorial set", at which point you're really not gaining much in a unified programming guide sense. I don't really know how you're going to solve this one.

--
"My life was more improved by a single use of [ecstasy] than someone's life is made worse by becoming a heroin addict." -- aphrael

Hyperlinks to building-block concepts (3.00 / 1) (#105)
by pin0cchio on Tue Sep 30, 2003 at 12:06:54 AM EST

Or perhaps force people to specify in the tutorial itself specifically which other tutorials should be required reading (with links or whatever) before attempting this one.

This would be easy: at the first appearance of "linked list" in an article describing an application of lists, insert a hyperlink to the article that explains linked lists. And in the article describing linked lists, insert a hyperlink to the article that describes pointers. That way, hyperlinks to building-block concepts plus the back button equal a depth-first search that should give the reader enough background. Wikipedia uses this approach to some extent; the only change would be a way to mark links to articles as links to "building blocks."

I believe that a collaborative textbook should discuss all the common paradigms without favoring one. For example, discuss how quicksort is defined in five lines of code in a functional language, and then explain how it works in a more traditional procedural language.


lj65
[ Parent ]
Another choice of software... (4.75 / 8) (#39)
by gusnz on Sun Sep 28, 2003 at 05:13:24 AM EST

I see Wikis have already been mentioned, but an alternative is the Everything Engine, which runs Everything2 (a user-written encyclopaedia) and Perl Monks (a Perl community and reference site) amongst others.

For anyone unfamiliar with it, the Everything Engine is an open-source content management system of sorts, with "nodes" that users create, interlink and vote upon. Each node also has a list of links to other related nodes, so you can get your web-of-knowledge thing going.


[ JavaScript / DHTML menu, popup tooltip, scrollbar scripts... ]

Linked Lists vs C (4.28 / 7) (#40)
by bugmaster on Sun Sep 28, 2003 at 05:47:21 AM EST

I believe that the worst possible choice of teaching methods you can make is to lump C and linked lists together into one category. These are two completely separate concepts; they even operate on different levels of abstraction. C is a low-level (well, relatively) programming language, whereas linked lists are a high-level data abstraction. You can implement linked lists in any language, or in pseudocode really. If you know how linked lists (and other algorithms and data structures) work, you can pick up any programming language pretty quickly. If you just know C, you still don't know how to program at all.

In fact, I think the reason we have so many incompetent programmers nowadays is because of teaching methods like these. People don't learn enough about data abstraction and algorithms, and, as the result, all they know is how to copy-paste code together.

Sorry, I know it's offtopic, I just wanted to get that off my chest...
>|<*:=

Linked lists (4.25 / 4) (#57)
by Peaker on Sun Sep 28, 2003 at 06:44:54 PM EST

are not high-level data abstractions at all.

They are data structures. Data structures define the low-level interpretation of memory, and do not define high-level abstractions. In fact, they are very concrete.

A prepared List Class in some language may provide the high-level data abstraction above a linked list, but that's a separate matter.

[ Parent ]

abstraction (4.60 / 5) (#71)
by kubalaa on Mon Sep 29, 2003 at 12:25:48 AM EST

I think the parent was referring to recursively-defined lists, also called singly-linked lists, or conses, or whatever. Linked lists is this sense are high-level and abstract in the same way that mathematics is high-level and abstract.

[ Parent ]
Perhaps instead . . . (4.33 / 3) (#62)
by ZorbaTHut on Sun Sep 28, 2003 at 07:48:17 PM EST

. . . you would have a few different starting paths. A few "programming" paths for various low-level languages - C, C++, Java, etc - and then a "computer science" path that goes into data structures in general. There would be a few lessons that had prerequisites in both of those, and these would be things like Linked List in C, Linked List in Java, and so forth, just to show how to implement them.

[ Parent ]
Re: Perhaps instead . . . (4.50 / 4) (#69)
by bugmaster on Sun Sep 28, 2003 at 11:06:17 PM EST

Basically, I think it's actively harmful to learn some specific programming language without learning what programming is all about, first. Linked lists, and simpler things like conditionals, loops and functions form the basis of all programming languages (well, except maybe PROLOG). If a person learns a programming language without understanding these concepts, he will just develop a habit of copy-pasting script fragments together: "if you want to make a program that rolls 2d6, take the 1d6 program and replace 1 with 2 on line 39". That's just wrong.
>|<*:=
[ Parent ]
Graph Models, Assesment and Cognitive Dissonance (4.45 / 11) (#43)
by voxol on Sun Sep 28, 2003 at 08:11:44 AM EST

It has already been mentioned how the knowledge base of a learning system is better modeled by a graph than a tree.

There is a paper (Aimeur et al, 2000) that describes a way in which capabilities can be mapped in terms of prerequisites and then lessons are modelled as gaining groups of those prerequisites, thereby setting you up for a new set of lessons. It also brings up the notion of different types of links: obligatory, optional, strong and weak contributions. The optimist in me says something like E2's hard/soft links mixed up with the described system might work.

The pessimist in me says that assessment is too necessary to miss out; but also too important and too specialised to provide in such a collaboratively built resource. One option is an assessment builder. Exam papers tend to consist of a set number of questions. Having someone describing these 'meta-questions' somehow a program could produce any length of assessment on demand. The system would naturally be limited in the type of questions that could be asked at first, but systems are being developed all the time.

Not directly linked to this, the Aimeur et al also describe the notion of 'cognitive dissonance' in learning. Cognitive dissonance is a state of mind where one experiences something that does not fit in with the knowledge obtained up till that moment. The brain is then forced to resolve the dissonance by either accepting new concepts and ditching the old or maintaining the old concepts and dismissing the new. The notion being that a learning system could utilise this by sneaking in false information and forcing the student to think on the subject. One advantage of the proposed system is that with the possibility of flawed information in the system the user/student is forced to evaluate the environment.

My own belief is that such a system could be built but finding willing contributors would be the difficulty. Manufacturers of electronic whiteboards have been trying to get teachers to create resources to share between each other in the UK and to my knowledge the uptake has been minimal.

  • Aimeur, E., Frasson, C., & Dufort, H. (2000) Cooperative Learning Strategies for Intelligent Tutoring. Applied Artificial Intelligence, 14 (5): 465-489.)


The problem of contributors (4.75 / 4) (#52)
by minerboy on Sun Sep 28, 2003 at 12:40:16 PM EST

Could be overcome by repurposing material already available on the web, and available as public domain, or through fair use. For example, there is a huge amount of information available through various government organizations. For example - and to be non-US centric, The UK's national grid for learning There are quite a few resources available, from NPR to merlot, etc. etc. - rather than creating content, the site would organize available content into a program of study, and provide feedback on the utillity of each resource

The assessment problem seems stickier - as well as some type of awknowledgement for users that improve their abillities. Maybe that is best left to traditional educational institutions, Though this is why they have the strangle hold on education.



[ Parent ]
Stick, what are you doing out of the diary (2.36 / 11) (#44)
by noogie on Sun Sep 28, 2003 at 09:06:13 AM EST

section?


*** ANONYMIZED BY THE EVIL KUROFIVEHIN MILITARY JUNTA ***
sounds familiar (2.44 / 9) (#46)
by crazycanuck on Sun Sep 28, 2003 at 09:47:50 AM EST

I guess we'll hear from you again in a year when you tell us your wife got tired of waiting around for you and finally contacted a lawyer herself...

Good Points... (3.42 / 7) (#47)
by Azmeen on Sun Sep 28, 2003 at 11:37:52 AM EST

However, I have to disagree with you on the unlimited editing period. Submissions should be able to mature and not live on forever as zombies.

This is specifically important when you traverse through significant version changes of the language itself... granted, this would be less of a problem on mature languages such as C, Cobol, and to a certain extent, Perl, for example.

My suggestion would be editing period should expire when there is a newer version/release of the language. This might seem meaningless to many, but the difference in code when moving from an older version to a newer one will be educational for newbies as well as seasoned veterans. It will also be a good way to perform a then vs. now analysis.




HTNet | Blings.info
I loved your article. (4.28 / 7) (#48)
by xutopia on Sun Sep 28, 2003 at 11:51:07 AM EST

Defining assumptions to me was the most important of them all.

Wonderful idea. Add chaos and stir. (4.87 / 8) (#49)
by danharan on Sun Sep 28, 2003 at 12:02:24 PM EST

This is an excellent post, and it has spawned some outstanding discussion. I will be so bold as to propose one question that could guide such a discussion: How can we design for your objective of a collaborative learning system with the fewest possible assumptions?

You're quite right to note that funding could be taken care of by advertising. However, I don't think it is really necessary that a single organization be responsible for everything (or a single editor- this will definitely require a wiki-type technology, as mentionned by other posters).

I also agree with many commenters that the tree-structure is not necessarily the best way to represent/describe knowledge. We do however, need some way to do this... and since that's often one of the most contentious areas in discussing such projects, I propose we just ignore it ;)

Or, more precisely, let competing 'narratives' compete: it's both impossible and useless to have a final arbiter on the web of what's a proper epistemology. If some want to build a tree, they can build trees... if other want webs, or maps, or yellow brick roads... the more the merrier.

How could tools build these representation? How can we have such tools and have knowledge scattered over many servers? I suggest each node have meta-descriptions with:

  • required knowledge
  • optional knowledge
  • links to required knowledge
  • links to optional knowledge
  • links to related nodes
  • links to more advanced nodes
If every node has at least one link, and most have more, engines could then crawl nodes, and represent information as they wish. In fact, I don't really like the first two items in that list, because they don't sufficiently separate knowledge from its representation... (can that actually be done? this might get some philosophers riled up!)

I hope this helps advance the idea... this sounds like a very important and exciting possibility!!

90% (4.50 / 4) (#101)
by ZorbaTHut on Mon Sep 29, 2003 at 08:21:34 PM EST

I think the problem I see with this (and with the idea of using Wiki or Everything2 or something similar, and with the entire idea of *not* having some high authoritarian Lead Editor) is the 90/10 rule. You can get 90% of the quality by doing 10% of the work. What you're suggesting here is 10% of the work. We don't have any firm way of linking nodes, we don't have a specific editor who's ultimately responsible for everything.

Basically, you're suggesting a wiki.

In fact, you *are* suggesting a wiki - and the thing about wikis is that we've already got them. And they haven't turned into collaborative learning systems. What makes you think this one will?

The problem with this philosophy is that you end up with a bunch of solutions that sort of solve the problem. The implementors finish the fun parts (i.e. writing a Wiki) and then ignore the not-fun parts (i.e. editing, screening offensive comments). And so it never really gets any farther. It's done, as far as they're concerned, and they never quite understand why it hasn't turned into a global repository for knowledge.

If this is really going to get off the ground, it's not going to be enough to just set up a site and say "Here! Have fun, people!" and leave it alone. It's going to have to be *much* more designed for the task at hand.

[ Parent ]

hmm... no, not a wiki (3.00 / 1) (#122)
by danharan on Wed Oct 01, 2003 at 11:14:55 PM EST

actually, anyone could put up a node, just like we put up web pages- whether or not it is in wiki format or not could be left up to contributors.

You're absolutely right to state that only takes care of 10% of the work though.

The other 90% - aggregators, search engines and/or mapping engines to let you navigate through all those nodes, might be a difficult undertaking, and it's not clear right now how we could do that. If we try to design this completely too soon, we WILL kill the project: there are too many assumptions needed to do the whole thing at once, and inevitably some will be way off base.

In this kind of situation, it makes more sense to me to figure out protocols and conventions, rather than how all the software will work.

[ Parent ]

Great Article! (2.09 / 11) (#50)
by Big Sexxy Joe on Sun Sep 28, 2003 at 12:03:08 PM EST

This is a brilliant troll.  I can't believe how many bites you're getting.  I tip my hat to you, sir.

I'm like Jesus, only better.
Democracy Now! - your daily, uncensored, corporate-free grassroots news hour
Great idea (4.66 / 6) (#51)
by ennui on Sun Sep 28, 2003 at 12:15:11 PM EST

Somebody taught me about Ben Jonson, I think. "Very few men are wise by their own council, or learned by their own teaching. For he that was only taught by himself, had a fool for a master."

Who needs an encyclopedia, after all, if you have everything2.

I wish you hadn't used the world "collaborative," it really makes me wince on this site. I read the title and thought "great, a copycat scheme/scam."

"You can get a lot more done with a kind word and a gun, than with a kind word alone." -- Al Capone

Another Ben Jonson quote (4.50 / 4) (#86)
by epepke on Mon Sep 29, 2003 at 11:11:24 AM EST

Only a fool writes, except for money.

Voluntary, web-based stuff is fine (hey, K5!) but it tends to hit an upper bound of complexity rather quickly. Unless you stick with academics, who are, in a sense, paid and given job security (in theory, at least) so that they can do this kind of thing on the side, who's going to write the advanced topics?


The truth may be out there, but lies are inside your head.--Terry Pratchett


[ Parent ]
In the Real World ... (5.00 / 3) (#99)
by suquux on Mon Sep 29, 2003 at 05:56:30 PM EST

Unemployed academics, practicing.

CC.
All that we C or Scheme ...
[ Parent ]

Learning objects and ARIADNE (5.00 / 7) (#54)
by IriseLenoir on Sun Sep 28, 2003 at 02:31:55 PM EST

I'm saddened to see no one else replied to this post or commented on learning objects elsewhere... the post is actually the sole reason I voted the article up... There aren't really any new ideas in the article, but the mentioned project is so exciting!

Everyone should go read at least the first part of the first document on the linked site : Connecting learning objects to instructional design theory: A definition, a metaphor, and a taxonomy. This is some truly amazing stuff, people!

I'll paste the two definitions, hopefully giving you an idea of the concepts involved and grasping your attention:

A learning object is "any digital resource that can be reused to support learning". It's basically object-oriented authoring, taking templates to a whole new level and allowing, at first human, eventually dynamic, mixing of small learning resources into complete specific lessons that can take many variables into account, most importantly the knowledge of the user, and eventually be an answer to the "assumed knowledge" problem. It also means that concepts common to many disciplines can be taught once and reused infinitely, blended with other objects about how they apply to a specific subject. Do not think of them as merely text, they can be anything: graphs, audio, video, etc.

"[I]nstructional design theories are design oriented, they describe methods of instruction and the situations in which those methods should be used, the methods can be broken into simpler component methods, and the methods are probabilistic."

Forget Scoop (which is not at all fit for the job and would require an almost complete rewrite)! A sister project (and probably more interesting and close to the author's goal) is the European ARIADNE. They define it as "A European Association open to the World, for Knowledge Sharing and Reuse, E-Learning for all, International Cooperation in Teaching, Serving the Learning Citizen." Part of the project is to develop GPL tools* for both authoring and learning, using PHP/Apache/MySQL (PostgreSQL supported). Tutorials, programming references, downloads and a tour of the software are provided on the linked site. The tour presents it more as a general web site authoring tool (which it can be used for, obviously), but there are tutorials on creating web applications that would certainly meet the project's requirements. The software already seems pretty mature.

Sure, authoring (quality content using learning objects) for such a system might require some basic knowledge of pedagogy/instructional design theory, but with professional managers to help divide the work, give a hand to authors and review/adapt content a bit for the system, such knowledge could be reduced to a widely accessible minimum. Many universities fund these projects and I'm sure some would be willing to at least provide some teachers' time to help.

Also, as minerboy (again! I love that guy!) mentions, there is a vast amount of quality, public domain or legally distributable information available on the web or that can be digitalised, which could be reused and adapted to such a system to save time and ensure that the project starts with very high quality standards and encourage contributors to follow in this trend.

Come on now, I want some quality brainstorming, here!

* The licence adds a clause that allows content created with, and that require ARIADNE to run, to be distributed under any licence, as long as distributed separately.


"liberty is the mother of order, not its daughter" - Pierre-Joseph Proudhon

Multicentric framework (4.33 / 3) (#75)
by zhiwenchong on Mon Sep 29, 2003 at 02:17:21 AM EST

You might want to look at this:
Multicentric framework for managing information
Lateral  Thinking revisited

Software to download  too.

[ Parent ]

doubts bout ariadne (3.80 / 5) (#76)
by dimaq on Mon Sep 29, 2003 at 05:20:43 AM EST

my understanding is that ariadne is one of them large projects funded by eu something something, in which case, nothing good will probably ever come out of it.

the difference in the proposed thingy AFAIU is that 1) it's 'grassroots'-powered, and 2) it's primary goal is to create a knowledge or learning thing bank

1.
[large science project funding inefficiency rant snipped]

2.
you can keep polishing a framework forever and try to push your ideas into it, but it doesn't count for nothing as long as there's no data in it.

right?

[ Parent ]

doubts legitimate - your view extreme (3.00 / 1) (#108)
by IriseLenoir on Tue Sep 30, 2003 at 03:47:02 AM EST

"you can keep polishing a framework forever and try to push your ideas into it, but it doesn't count for nothing as long as there's no data in it."

Yes, yes, of course (well, no it does count, just that it's not useful to the final user). My suggestion was exactly this - use their tools and ideas to create what we want with. I agree with a rational variant of your extreme point of view - that ARIADNE most probably never will create anything close to what the author suggests in terms of quantity of data. But the software is GPL and ideas are free (as in speech), that's the important part.

I think that to say that there will never come any good from a project just because it's funded by a EU organisation is pretty stupid. Also note that the main organisation is by EU, but many European universities fund independent projects based on it.

You have some valid points, but come on! Can't believe you have a perfect score after 7 moderations. Gave you a 2 for balance ;-)

"liberty is the mother of order, not its daughter" - Pierre-Joseph Proudhon
[ Parent ]

I understand your skepticism, but... (none / 0) (#136)
by Erik Duval on Sun Nov 02, 2003 at 05:02:12 PM EST

I do understand that people are skeptical about EU funded projects, but one should notice that ARIADNE actually grew out of a certain feeling of frustration with the way most such European projects go. We explicitly want to avoid the continuous disappearance of project results, that so typically happens as the project members go chase after another round of funding

In fact, we have been operational for more than 7 years... ARIADNE is now a self-sustained Foundation, funded by membership fees. We are open to collaborate with the rest of the world: drop me a line if you want to know more...

[ Parent ]

the hard part (4.71 / 7) (#56)
by Work on Sun Sep 28, 2003 at 05:13:44 PM EST

a decent idea, as you're basically talking about taking college or university style of learning and putting it up for a more free collaborative effort.

However, the problem is difficult and abstract concepts. I'm sure as a self-taught programmer, you've never investigated deep into discrete math or such things as induction proofs. They're difficult and their usefulness often isn't immediate, but its one of those things that sticks in your mind and makes you a more efficient and competant software engineer. Mainly because with proofs you can make your software more robust and less prone to bugs. And often better written in general. Less spaghetti code. And once you've done enough of them, you unconsciously introduce elements into your everyday code.

Anyway, back to my main point. Things like induction proofs can be very abstract and difficult to grasp. A university student *has* to learn them in order to pass his tests, and thus his class in order to go onto the next level. Most of us probably would choose to never study automata theory and the like if we had the choice.

With this system once a person comes across a difficult concept to quickly grasp, they are likely to skip it and move on. However this may mean they're underqualified to move on, which propagates in their further lessons.

Au contraire (4.33 / 3) (#66)
by sctfn on Sun Sep 28, 2003 at 09:22:59 PM EST

I really have to dispute this point. As a self-taught programmer myself, I do believe I know plenty about things like induction proofs and automata theory. And, in fact, I found them quite entertaining to learn. However, what I cannot say is that they regularly conciously impact my programming. At the moment I'm working on yet another lisp as a little toy language, and I can safely say that, as I prepare to use it to teach myself assembly, I really am not using that much formal foundations of math/computer science in my work. Just my 2c.

-Scott

[ Parent ]

assembler (4.60 / 5) (#68)
by Work on Sun Sep 28, 2003 at 10:27:53 PM EST

if you plan on going past asm and into how hardware actually works (which to grasp assembler, you really need to), automata theory comes in handy. They both use finite state machines and pattern recognition. Actually thats all digital hardware is for the most part - FSM's looking for certain sequences.

Yes, thats oversimplifying things a tad.

interestingly, most programming is taking a non-deterministic problem and turning it into a deterministic one the computer can understand. Theres a bunch of good algorithms out there for doing this, but many programmers don't know anything about it and just kind of do it haphhazardly without realizing it. Which is why software tends to be so buggy and hardware tends to not be. Hardware engineers know their FSM and state minimization algorithms implicitly through their work and how to handle EVERY case (at least within reason, some big systems are difficult to do this for), but software coders don't and shoot themselves in the foot unintentionally, even in small and simple problems.

[ Parent ]

Academia vs. School of Hard Knocks (4.50 / 4) (#85)
by epepke on Mon Sep 29, 2003 at 11:00:41 AM EST

Ah, the old battle. Known to generate as much heat as vi vs. Emacs. Yes, of course there are plenty of university graduates who can't program worth a lick, and there are plenty of self-taught individuals who are excellent. Aside: don't they teach basic sequences and series and inductive proofs in high school any more?

But anyway, I agree. The value of formal CS (even if learned in a non-university setting) is in how it gets under your skin and gives you an automatic "ding-dong" response. Nothing grows spiky hairs faster than a quick-and-dirty parser, unless it's built on solid theory.

I started off as a self-taught programmer and did all the things that the ilk tends to do: LISP interpreters and threaded interpreted languages and the like. When I became of age and decided to go for a mathematics degree with a concentration in CS, I hated it for seeming so impractical. Now, many years later, I'm incredibly glad I had things like graph theory and Karnaugh maps and Turing machines hammered into my skull.


The truth may be out there, but lies are inside your head.--Terry Pratchett


[ Parent ]
karnaugh maps (4.50 / 4) (#88)
by Work on Mon Sep 29, 2003 at 11:24:49 AM EST

ugh. i hope i never see another 5 variable kmap. transition tables, state diagrams... I once took this mindraper of an EE class which started with the basic AND/OR operations and ended with the final project of building an actual calculator in hardware in just one semester. I put something like 20 hours or more a week into that class.

[ Parent ]
Linked Lists (3.66 / 6) (#58)
by Peaker on Sun Sep 28, 2003 at 06:48:44 PM EST

Linked lists are structures used to hold data in programming languages. They are mainly used in C programming.

Linked lists are data structures used to hold data (not specifically in programming languages).

I could understand "They are mainly used in Lisp programming" (where almost all data is built of connected CONS pairs that form linked lists), but even that would discredit the generic data structures that are useful for many algorithms.

C programmers, as all programmers, really should choose the data structure appropriate for their algorithm, and I don't see why C would form any more bias towards using a linked list than other languages.

Going a little bit further... (4.50 / 4) (#64)
by cribeiro on Sun Sep 28, 2003 at 08:47:27 PM EST

I think that the author intended to simplify the writing - to make it shorter as not to obscure the main point of the article. In the process, the footnote ended up being simply wrong, as you point out. First, because one doesn't hold data in programming languages - actual programs do hold data on memory... and the reference to C programming was unfortunate, in my opinion.

I understood it as this:

Linked lists are structures used to hold data on computer programs. While some languages provide direct support for lists, other languages (such as C) force the programmer to write code that manage them. Linked lists are simple but powerful, and because of this, they are commonly used to introduce programming concepts such as pointers.



[ Parent ]

Where is Paul Allen when you need him? (4.44 / 9) (#60)
by CoolName on Sun Sep 28, 2003 at 07:15:23 PM EST

This might work in programming but in the humanities exactly what the subject is is a matter of debate. The decision to teach a certain history for example would almost have to be a political decision rather rather than an educative one. There is always a danger of putting out what is really propaganda under the guise of education when there is a monolithic source of information. Here is an idea for the humanities though. Perhaps some Internet billionaire could pay for Internet subcriptions for essentially the whole world , make the magazines free on the Internet, to say 15 or 20 high quality magazines such as Nature, Science, Foregin Affairs perhaps some top medical journals etc. The debate in this country would be substantially raised. Where is Paul Allen when you need him?

"What does your conscience say? -- 'You shall become the person you are.'" Friedrich Nietzsche


Multiple viewpoints (4.33 / 3) (#102)
by marsbard on Mon Sep 29, 2003 at 10:59:09 PM EST

You could actually improve on the current teaching system by presenting multiple differing versions of the same historical event and allowing students to make their own critical distinctions between them, based on other available versions of related events.

Attaching a discussion forum to lessons might help get around the propaganda problem - at least propaganda would attract a lot of attention from its opponents.



[ Parent ]
Sounds like (4.00 / 5) (#63)
by ZorbaTHut on Sun Sep 28, 2003 at 07:51:17 PM EST

a great idea. When is someone going to get around to programming it? I'll write lessons once the site's up.

How to you keep people motivated? (4.71 / 7) (#65)
by cribeiro on Sun Sep 28, 2003 at 09:15:38 PM EST

First of all, a short intro. I wrote a diary entry on a related topic a couple of days ago (yes, I know you're aware of it :-). In short, I'm thinking about starting study groups to explore alternative ideas.

Study groups are far from being a new idea, but it seems that they can be much better explored not only as a learning tool, but also as an effective social tool, specially for things that have potential to change the way we live our daily lifes. But why don't people join spontaneously to study something together? There are several problems, but the biggest one (in my opinion) is how to keep everyone motivated.

Motivation is the difference between a succesful study group and a abandoned one. It's also the difference between a successful website and a abandoned one. So, the big question is, how do you intend to keep people motivate?

There are several ways to tackle this problem, but I'm not aware of any definitive experiment that shows how it can be done successfully. Some people seem to believe that it is just a matter of having good people in the group - I mean, people that really understand the topic so they can push the group. Others believe that the group should be balanced. But in the end, both alternatives (and several others) are based on the assumption that a good group can manage and keep itself alive. I'm a little skeptical about it.

I think that, even with the best people in the group, it's very difficult to guarantee that everyone will be involved to the same level. So it's important to develop some external mechanism to foment participation. It does not need to be money; but some type of incentive or economic pressure is important. People are competitive by nature, and it is important to have something at stake.

In short, I think that motivation is the biggest obstacle for a successful implementation of your idea. Analyzing the reasons behind the success of some websites should be useful. If you solve it, you're almost done - the community will do the rest.

Open Source of some kind might help (4.33 / 3) (#73)
by chanio on Mon Sep 29, 2003 at 01:01:25 AM EST

I have thought that countries that have free universities should use the students to get some work done as a test that they are learning.

Then, what would be the need of having examinations?

But everybody would be interested in seing that what they have studied has become something useful for their comunity.

If that idea could be better developed, we might see the day that every job could be another school.

And the reward might then not be money but ending every year with more wisdom or knowledge...

The most interesting part of a group of study is that one can learn to think with others, besides the subject of the study, don't you think? That is the main reward.


________________
Farenheit Binman:
This worlds culture is throwing away-burning thousands of useful concepts because they don't fit in their commercial frame.
My chance of becoming intelligent!
[ Parent ]

Social aspects of study groups (3.00 / 1) (#114)
by cribeiro on Tue Sep 30, 2003 at 05:50:32 PM EST

If I understood what you've said... I agree that the social aspects of studying with a group of people are very important. When you study alone, you may end up stuck with a few of your own pre-conceived ideas, and not to listen to anything else. In a group, you have to learn to listen to other people; to be patient when someone does not understand what you are saying... and finally, to understand that you may be wrong, and someone else right - even if you don't like the chap. It takes a lot of maturity to reach this point.

[ Parent ]
motivation (none / 0) (#132)
by chanio on Sat Oct 11, 2003 at 07:05:56 PM EST

You are right. And that is learning :).

I wouldn't mean learning anything that is not related to me as human.

There are some motivation examples in individuals that do things without a material reward: like olimpic games, even climbing high peaks, exploring unknown regions, etc.! Finally, these individuals join happily in a competition, exchanging experiences, sharing their fears, helping each other to achieve their own goals...

There is a sort of satisfaccion when one of these individuals reach their goal.

Even if I am not able to do it, I am happy to know that someone that shares my main interests finally does it.
________________
Farenheit Binman:
This worlds culture is throwing away-burning thousands of useful concepts because they don't fit in their commercial frame.
My chance of becoming intelligent!
[ Parent ]

Creativity (4.50 / 4) (#90)
by paranoid on Mon Sep 29, 2003 at 12:03:08 PM EST

I replied to your diary entry, but I will do it here in a bit more details.

You can't motivate people with something external to them, unless you have a million dollars and are ready to pay it to people for studying. :)

A better way is to make them motivated (by letting them motivate themselves). Let them create something on the basis of your learning. Start a project or something. Write a scientific paper together, build something, code something, etc. That would create the motivation to study, because suddenly all participants need the new knowledge to proceed with the project.

[ Parent ]

Motivation and capitalism (3.00 / 1) (#112)
by cribeiro on Tue Sep 30, 2003 at 04:50:24 PM EST

I agree with you to some extent. However, I think that we have to better understand how to apply current economic thinking - including a few capitalist principles. It's not a matter of being a liberal or right-wing radical - I'm very far from it. But it is a matter of accepting that you can't ignore the market forces, and that it does have a lot to do with the way people make choices and decide where to spend scarce resources, time included.

Looking at other things may help to explain what I mean. Some people are advocating the use of market mechanisms as a way to encourage companies to limit pollution. Special boards are being set up to allow companies to trade polluents as credits and debits. This is a clear advance in the way ecologists think; instead of utopian thinking, this is a practical application of the market mechanisms to tackle ecologic issues.

What this does have to do with study groups or collaborative learning? Well, the academia can't keep going thinking that they are not part of the market. Even pure science has a value; the problem is how to measure this value. Alternative learning systems have to take this into account too. Students are going to invest time. Authors too invest time writing. How do you keep everyone motivated? it is important to understand how does this market works; how does it makes its choices; and how does it assign values to things. This is a problem with psychologic and economic aspects. Failing to understand the economic side will undermine the overall understanding of the problem.

[ Parent ]

Make learning a game. (3.00 / 1) (#119)
by xicubed on Wed Oct 01, 2003 at 02:09:19 PM EST

I thought I'd start this comment by comparing the amount of time people spend playing online games vs studying. I went searching and grew bored but it must be some huge number. On this site which collects information about games a guy wrote about what my point was going to be anyhow so I'll link to it. Somehow harness the captivation games provide and use it towards learning. link here

[ Parent ]
I have this problem (3.75 / 8) (#70)
by auraslip on Sun Sep 28, 2003 at 11:29:31 PM EST

I'm enrolled in a local community college. If I do well in all the classes I can transfer to a university.
The problem, as you put it, is that %90 of what I sit through in class is redundant, if not completly shit obvious. I spend a good many hours wasted learning nothing.
If you fall behind your level at all, you have to work to catch up.
And I am not alone, my class mates are not stupid. I know they are bored too. If community college doesn't kill off any will for education nothing will be able too.

Often it also happens that I'm interested in learning something, lets say neuroscience. I can go to the libaray and get a book and learn a great deal on the subject, but then in higher education I'll have to sit through boring redundant shit.
___-___

Ideally (4.60 / 5) (#80)
by michaelp on Mon Sep 29, 2003 at 06:44:51 AM EST

you ought to be able to learn on your own, and take some sort of standard test that establishes whether you actually know anything & gives you units if you pass. Western Govenors U is the only accredited college doing this afaik.

Interestingly, the work of Piaget and his students would suggest that that is the best way for most people to learn: get interested in a problem, learn all you can about it, try to figure it out on your own, etc. rather than sitting through lectures with a hundred other bored undergrads. For economic & cultural reasons we mostly teach in the latter form though science has repeatedly shown us it is deeply flawed.

A big part of the problem is developing fair and usable assessments that actually test higher order thinking, so far this has proven a challenge and most standard tests mostly test rote memorization rather than higher learning while having experts review and grade your work is very expensive.

However, new warez are being developed that let computers grade essay tests, math problems, etc. and so it may be that more Us will start to focus on finding out what students know and helping them to learn rather than on filling seats with butts.

"Every gun that is made, every warship launched, every rocket fired, signifies in the final sense a theft from those who hunger and are not fed, those who are cold and are not clothed."

[ Parent ]

I was going to mention the technology factor (4.33 / 3) (#100)
by auraslip on Mon Sep 29, 2003 at 06:24:03 PM EST

But even if you can learn at your own rate with the computer, you don't work with people. And that is often more important than what your actually learning.
___-___
[ Parent ]
Well having c ompleted numerous (3.00 / 1) (#107)
by michaelp on Tue Sep 30, 2003 at 01:52:24 AM EST

sucessful group projects at work, myself, seems I ought to be able to test out of that requirement as well:-).

PS, one of the recent additions to the open source learning management software I mentioned in my other comment on this article (moodle.org) is a collaborative project module which enables online groups to be assigned a project and work on it collectively (via hand versioning or wiki) & grade each other's work.

"Every gun that is made, every warship launched, every rocket fired, signifies in the final sense a theft from those who hunger and are not fed, those who are cold and are not clothed."

[ Parent ]

School can be so boring (3.80 / 5) (#81)
by Viliam Bur on Mon Sep 29, 2003 at 07:05:46 AM EST

I left university this year, because the ratio KNOWLEDGE_GAINED / TIME_SPENT was terrible. First years I believed that perhaps later it will be better. No it did not.

Most what teachers did was slooooowly saying what was written in some book. After spending 90 minutes in classroom, all you had was 2 or 3 sentences that made sense. Add some useless stuff, just to make enough questions for exams.

Example (I studied psychology): In the 3rd grade we have learned about Sigmund Freud. When was he born. Where did he work. What "keywords" did he use. And we had an exam with these questions. But most of us did not know what those keywords mean -- we never had a discussion on it. When asked "Who used the term 'Oedipus complex'?" you had to answer "Sigmund Freud". But 9 of 10 students did not know what 'Oedipus complex' means. I have asked the teacher, but she was not able to explain it... she just told me to find a book and read about it. She had not time to explain this. However, she had spent 90 minutes telling us what year and in what city was the guy born, where did he study, and where did he work. Surely that is the important stuff which is supposed to give me a deep insight into human psyche.

Such teachers could be so easily replaced by a web. I wish they were. Soon. So I am looking forward to any such project.

[ Parent ]

a lesson from history: let it flow... (4.50 / 6) (#72)
by chanio on Mon Sep 29, 2003 at 12:43:42 AM EST

By experience we should know that any natural process like learning shouldn't be considered but from the close need of it.

WWW is very new, but in a close time, this need is going to arrise as something imperious to solve.

That would be the exact moment of starting compiling all the partially developed ideas.

Like for example wiki forums. They have almost reached your idea. Every new concept that we want to acquire,to be understood, needs a lot of concepts and parallel thinking. It might be like the structure of a tree upside down where every branch joins with others until reaching to our actual situation.

That is: one cannot anticipate what is required to understand any new concept. But everybody can check what he doesn't really understand (just being honest with himself).

A well structured wiki might let you learn everything and later return to your previous situation of learning a new concept.

There are a lot of teachers that are anticipating the future of learning. I think that they are idealizing an ancient time when learning had no divisions. It was just an interesting dialog that let everybody explore new concepts and learn at the same time lots of subjects: psicology, mathematics, biology, literature, etc.

But we shouldn't plan how it is going to be done until the need becomes evident for everybody.

I think that in the future the www is going to be like a universal mind. And that is going to have even techno-psicologysts to repair it from time to time :) .


________________
Farenheit Binman:
This worlds culture is throwing away-burning thousands of useful concepts because they don't fit in their commercial frame.
My chance of becoming intelligent!

Right on! (4.33 / 6) (#78)
by dimaq on Mon Sep 29, 2003 at 05:54:27 AM EST

go look for existing projects on the net, figure out how your idea is better, make sure it is way better, start coding, set up a project on sf.net...,

just please, please, please, don't focus too much on framework yet - rather start gathering data, until you see how much framework is really needed.

do keep k5 updated on your work, I'll gladly contribute.

my random ideas if you need em -
1. it's probably some sort of a graph than a tree.
2. each 'article' (e.g. linked lists) should come in different variations, or perhaps optional sections - one might want to check O()'s on operations on linked lists, another a proper way to implement em in language X, in which case, I'd see the presentation as a table with collapsible columns, say theory on the left, C examples on the right, and ML, Java and general concurrency issues collapsed.
3. the way I see editing work is that Another Random User creates an alternative version of a particular topic, or part thereof, and then one of original author, generic site editor, actuall readers decide which version is better.
4. however distributed the site is, I think it will still need at least semi-permanent editor that kill garbage articles, correct factual errors or coding style, and so on. Such person is sort of like last authority in the conflicts between authors and stuff. In short, users cannot be relied on to author articles too well, or to use the framework too well.

Speaking of an OSS learning management system (4.60 / 5) (#79)
by michaelp on Mon Sep 29, 2003 at 06:15:17 AM EST

to provide the framework for this curriculum (which sounds like a great idea) I'd suggest Martin Dougiamas' open source, PHP/MySQL learning manangement system, which has such things as quizzes, assignment posting, gradebooks, peer review, course creators separate from site admins, active community of users constantly adding new features and a commited lead developer expert in online pedagogy (who's also a decent programmer:-).

In fact at my dayjob we've been testing/reviewing open source LMSs as alternatives to our $70,000 LMS and Martin's product, Moodle is about the only Open source LMS we've found to be feature complete (compared to the high end commercial systems like Blackboad and WebCT).

Check it out


"Every gun that is made, every warship launched, every rocket fired, signifies in the final sense a theft from those who hunger and are not fed, those who are cold and are not clothed."

You forgot.... (4.33 / 6) (#82)
by moonpolysoft on Mon Sep 29, 2003 at 08:31:21 AM EST

... that MIT now has much of their class material online. I would venture to say that the online class materials do a good job of covering your points.



examples, examples, examples (4.66 / 6) (#83)
by harrystottle on Mon Sep 29, 2003 at 09:06:52 AM EST

speaking as another self taught programmer the most valuable and informative education I've experienced has been through crawling over other people's work and seeing how they've used a particular command or syntax. Include the option, therefore, for users and editors to add any and every example they think is relevant to a given lesson... (with the editors having the final say on whether it really is relevant)

Mostly harmless
While I agree that.. (3.16 / 6) (#84)
by Sesquipundalian on Mon Sep 29, 2003 at 10:48:47 AM EST

the existing channel based info-buffet that we enjoy as bipedal hominids is absolutely fantastic! The radio stations, television broadcasts, video rental networks, gossip, web search engines, usenet servers, the domain name system, the FTTP_CDRW interfaces, the peer2peer as in beer, the messaging, the immersive and the interactive (up to and including war, actually). Oh the inexpressibly cardinous (numerous) set of all possible tounges of thought!

It also has some "settling down" to do.

Your Idea won't work. Information is too slippery to put in a tree. Even a tree with lovely nests as you seem to have worked out in excruciating detail here. Information'll always slide out and flop all over the ground. In fact, no matter what container you put it in, information always escapes away from you. Despite some people, who want information to be free so that free can become valid consideration in a contract, Information leaks. Out of the tree and into abstractions that come back and affect your original understanding. (As others seem to have noticed as well), Your article would be more usefully interpreted as an on-line information resource, user interface design, How and What to.. article.

If I can "embrace and extend" an idea from Greggory Benford::Sensorium; Take your five senses, or if you prefer Gardner, your sensory modal inputs.. plus all of the stuff you think automatically when you first get a stimulus, you know.. like all of your knee jerk reactions. Add that to all of your ongoing but unconfirmed suspicions. Throw all of your current propositional(the ones that can be sensibly stated) and nonpropositional (the ones that can't) beliefs on to the pile and finally add your habits, passions, inklings and inclinations. and you have what I will now refer to as your sensorium. This is the part of you that brings data to the rest of your mind (pun::can you guess which part "the rest" is?). Do not be distracted by the fact that this is also the internal shape of your life.

Your sensorium improves in both resolution and accuracy; with training (take training in the broad sense here).

My point is that a well developed sensorium is the proper user interface for all_of_knowledge, not a tree. So to E'n_E some wild west Darwin, come down from the tree, train yer-self up a right-smart-sensorium ta use, and dig in to some of the fruit that your descendants have made to share with you.

By the way; Really cool article!



Did you know that gullible is not actually an english word?
Integrating content (4.50 / 4) (#87)
by paranoid on Mon Sep 29, 2003 at 11:24:32 AM EST

Here is an idea on how you can integrate different pieces of content. Instead of using hyperlinks, one can think of the articles and paragraphs as complex blocks. Taking your example of linked lists and C, you can have a short intro to C with one paragraph about linked lists. This paragraph should be connected with a more detailed explanation (say, one page) and even a longer article (say, ten pages). This way, it will be possible to preseve a linear structure (it might be better for a textbook) while allowing to customise the material.

When someone starts using such textbook, he can pass a short test that would determine what he needs to learn. If linked lists are necessary, the textbook will include the longer explanation by default.

This doesn't really help to prepare such materials, but once you have them, it might be helpful for personalisation of content.

You could have a diploma by now (3.75 / 4) (#89)
by johnnyfever on Mon Sep 29, 2003 at 11:49:20 AM EST

...from a technical school (you said you'd been figuring on this for 2 years.) Just thought I'd point that out :)

Infinite Monkey syndrome (4.33 / 6) (#91)
by Will Sargent on Mon Sep 29, 2003 at 12:24:29 PM EST

While this is a good idea in theory, I believe it falls victim to the Infinite Monkey syndrome. Basically put, there are just more monkeys out there than good coders, and some of them will jump on you and start posting crap. If you don't already have the experience to know the subject matter, then it may even sound convincing to you.

And for some reason, there's a lot of crap out there. The architect of C# wrote an article about the trouble with checked exceptions. Holub (normally a very good writer) apparently wrote about Getters and Setters being evil (well, not really, but it certainly looks like that at first glance). Greenspun... well, he's always had problems, but he's far worse than he used to be.

So without a paid, experienced editor, you're sunk. Of course, if you had a paid, experienced editor, you'd talk to him about his experience and not bother trying to weed out the crap.

Some people say better off buying a bunch of college textbooks, reading through them and discussing them in person with a group of your peers. But then you're in College.

But there's always going to be some information missing. It's the paradox of the missing link -- every link you fill in the middle creates two more to either side. There has been a tremendous flood of knowledge since the Internet, and things like WikiWikiWeb and Artima were very useful in the day (before they fell to the syndrome).

Or you can do what I've done -- write a website, talk about what I screwed up and what I'm trying. You don't get much information back, but it's a useful way of spreading knowledge, and I can point people to the website instead of waving my hands around and making diagrams on napkins.

Ultimately, the only experience you can trust is your own. You can figure out how good other people are by how much stuff they tell you to do works. If it doesn't work, fix it. :-)
----
I'm pickle. I'm stealing your pregnant.

Adaptive hypermedia, graphs, and other animals (4.50 / 6) (#92)
by ManuEfe on Mon Sep 29, 2003 at 12:31:20 PM EST

I'm new to kuro5hin, but as a newbie researcher in the field of web-based learning, I just couldn't resist a few comments. Kudos to the poster for bringing up this subject.

Now for some summing-up: You are proposing to re-do WebCT (or Blackboard, or TopClass) in an open-source fashion, and with multiple collaborating authors that submit peer-reviewed material to the course. A second interesting idea is to keep well-groomed "defining assumptions" for each module. This boils down to two issues: collaborative course authoring, and metadata on the nodes.

I would like to add something from my field of experience: adaptive hypermedia. The idea is to present a user only "relevant" parts of the course, by taking into account his "profile". Where the profile is both static (ie: what have you studied) and dynamic (what modules have you visited/liked/scored poorly). Plainly stated: get the course adapted to the student, and not the other way around. You can adapt it two ways: change the contents, or change the structure (as in suggest/disencourage hyperlinks). We do mainly the latter. No big deal?

It turns out that this /is/ a big deal, because of a few factors:

  • When the number of alternative paths gets high, and the complexity of the decision rules (ie.: what paths are recommended over others for a given profile) increases, the structure becomes a nightmare to author.
  • When the number of modules grows large, it is very hard to ensure that all paths leading to each path makes sense.
  • And finally, when updating contents, maintaining different versions of modules that are linked from different places with different requirements is not an easy task (similar to DLL hell).

I have recently finished a project that suggests using a graph instead of a tree to lessen problems #1 and #2. The big problem with graphs, as one comment stated, is that they tend to be much more difficult to create and manipulate than trees - allthough a few techniques do help out. If anybody has time to burn, check out the pdf (comments would be very apreciated). In my view, knowledge is a web much more than it is a tree - and as one poster put it, you need one hell of an editor when you edit something this complex.

Other features we are trying to add to our homegrown (and soon-to-be-GPLed) adaptive course system include collaborative tasks. The problem with motivation (as noted by a few posters) is very real: if you sit alone in front of your screen, clicking "next lesson" (*must avoid Simpson quote*) every now and then, you had better bring all the motivation you can muster. If you can get students to work with each other solving tasks, or simply see who's on the same problem and ask/offer help, courses would be less passive. Getting lessons that are actually interesting (adaptivity!) helps here, too.

As several posters have written, there are multiple efforts underway to standarize metadata-annotated content into "learning objects". SCORM (linked here) is a good source for more info. Most of these schemes use some kind of semantic markup (RDF or some apropos-XML) to make dependencies machine-parseable. But the topic of knowledge management and reuse is still far from solved - even with well annotated contents, you have to ensure that authors can find what they need and seamlessly adapt it into their courses.

Another challenge is what to do when everybody can contribute to the knowledge pool, and there are no pre-defined courses. You then need some kind of planning system that given a "learning goal" and the student's previous experience and/or actions, searches in the knowledge pool and arranges relevant modules into a well-knit sequence (or a series of alternative paths). Re-planning would be necessary if the student chooses to go for another route. An alternative is simply not to plan, simply keep an eye on the student and hint what the next "logical" step may be (or where to go to get more background or such).

One final comment: if the system is really large, server-side scripts such as PHP will probably not cut it. If modularity and maintainance is needed, something more structured, such as Java, may come in handy.

Ahem, I think this will be enough preaching for the day. Please don't flame my previous paragraph too much.



nice comments (3.50 / 4) (#95)
by minerboy on Mon Sep 29, 2003 at 02:03:05 PM EST

Before I start, I should point out that I have some radical views on web based education. I enjoyed reading your post, but I see a few things differently. Part of the beauty of such a system is that it breaks out of the mold of institutionally structured coursework. Imagine people learning for learnings sake (or even just to improve their employment condition). These "students" should have sufficient motivation, be able to adapt the learning resources for their own needs, and create useful collaborations on their own. I would hope that the system would not follow a collegiate model at all. I would also expect that the user / student would have some problem to apply the knowledge to before coming to use the resource. This may be along term goal, but might provide some immediate and direct application of the newly acquired knowledge.

Using adaptive hypermedia in such a project has some interesting possibilities - perhaps, if the material is adaptable rather than adaptive, someone could write some type of adaptive filtering plugin for a specific user group - for example, something aimed at highschool students. It might function in a manner similar to the pornoblockers (but hopefully more effective) This might solve the complexity problem.

Last, I reject Scorm whenever I can, because it is the standard for military training, and I prefer some of the other metadata standards. On the other hand, many IT professionals are embracing it because of $$$$$$$$ - there is lots of pork to be had in army training these days



[ Parent ]
LON-CAPA (3.50 / 4) (#97)
by Jerf on Mon Sep 29, 2003 at 02:43:48 PM EST

LON-CAPA does most of what you mention, and is planning on doing the rest (we don't really support adaptive paths yet because nobody has really wanted to use them yet... BTW if you think a graph structure makes a learning system hard to use, just wait until you have a graph structure that is constantly changing! We're still confident it's doable but the number of issues that arise is appreciable.). We already use a graph structure for directing the student. Multimedia content is already in the system. A highly-adaptable homework engine is in the system as well. (Check out the PDF documentation from the "Documentation" link on the home page.) The system is designed to support large networks of content.

Peer review isn't in per se (though we do have the ability for multiple people to work on the same resources), but will be sooner or later. (It turns out to be a lot harder then you might think.)

And it's already GPL'ed, and serving tens of thousands of students.

[ Parent ]

A proposal for writing fewer stupid proposals. (3.50 / 5) (#94)
by tkatchev on Mon Sep 29, 2003 at 12:51:04 PM EST

The only proposal I would care for.

   -- Signed, Lev Andropoff, cosmonaut.

Civilization Tech Tree (4.33 / 3) (#96)
by Rubyflame on Mon Sep 29, 2003 at 02:25:22 PM EST

In fact, I'm thinking that the perfect analogy for this system is the tech tree in Civilization. You could create a complete tree of lessons, with each one having a list of prerequisites. When you start out, you can check off all the lessons you already know. Then you could get lists of whatever lessons are available for which you have completed the prerequisites.

Tech trees take a root for granted (4.33 / 3) (#98)
by ManuEfe on Mon Sep 29, 2003 at 03:51:36 PM EST

In a course, that would be the "goal". And not everybody has the same goals when studying a course. Sometimes you simply want a quick freshner, other times you simply want to get as quick as possible to "Railroad", never mind what the real "goal" is supposed to be.

The idea of browsing the structure and placing a checkmark on everything you already know is nice, though.

I still think that a graph is better suited to represent the structure of a course - there's simply too many ways to look at things for a "one size fits all" approach.



[ Parent ]
free books (4.33 / 3) (#103)
by bcrowell on Mon Sep 29, 2003 at 11:17:41 PM EST

Here are some free books on computer science. I have to admit I'm biased in favor of real books written by authors who really sat down to write a book. Stuff like wikipedia is fine, but you don't have to dig very far in wikipedia to find out how unreliable it is. For instance, check out some of the articles like Zodiac and Astrology. I also like writing where the author's own individual voice comes through. For example, the O'Reilly "camel book" (on Perl) wouldn't be nearly as much fun if it didn't have Larry Wall's voice.

The Assayer - book reviews for the free-information renaissance

Excellent! (3.00 / 1) (#104)
by James Hepburn on Mon Sep 29, 2003 at 11:52:33 PM EST

A great resource that I wasn't aware of, thank you.

"Computer Science is embarrassed by the computer." -- Alan Perlis
[ Parent ]
Meta data + search engines (4.33 / 3) (#106)
by xasmx on Tue Sep 30, 2003 at 12:30:15 AM EST

I've experienced the same problem. It's relatively easy to extend your knowledge to a field close to your current knowledge. When you try to start understanding a field where you base knowledge is next to zero, the start can be pretty exhausting (usually not because the field is so hard, but because there is just such an information overflow, that it's next to impossible to get a good 'big picture' of the basis).

I would see the ideal solution as a combination of meta data, and search engines. Meta data needed for such a thing (as mentioned already by many others), would prerequisites (and skill level on them), skills 'gained' from the text (and an indepthness those are gone into), etc.. To not be 'total chaos' of key words, the pre/post classifications should be codes refering to a standardized tree of 'concepts' (which would surely be much easier task to create than a full tree trying to teach those concepts).

A search engine would build graphs based on the pre and post elements, and allow the person to then choose learning routes him/herself. Some of the reasons for such a meta-data/search engine approach, versus some static tree/graph approach: More scalable. Anyone can participate. Experts often totally disagree with each other, which I think would make creating a static moderated structure hard, without choosing sides.

Some of the problems would include: Abuse by web-site advertisers. Prevention of the system getting over flooded with average texts.

Perhaps some kind of a centralized rating system might help here, where a centralized database would keep average score given by readers to the articles for several different viewpoints, such as qualify of writing, relevance, how well it fits prerequisites, etc.. Of course the rating system could be abused too, but I could see ways on how this might be made statistically insignificant using some heuristics and weights.


Semantic web and ant trails (3.00 / 1) (#113)
by ManuEfe on Tue Sep 30, 2003 at 05:06:23 PM EST

If the semantic web effort takes off (have a look at the w3c), this kind of searching will be possible. Until then, search engines have very little clues on what is potential course content and what is just (say) a book review.

Once you can scour the web for contents, and these contents are self-describing (as in your post), it would be easy to plan/build courses on the fly.

An interesting way of rating would be to use something similar to "ant trails": reinforce those paths that are most frequently used, and fade others . As people are not as uniform as ants, you could input a profile - and the system would then give more weight to people with similar profiles.



[ Parent ]

It would be very nice... (3.00 / 2) (#109)
by YagoBlade on Tue Sep 30, 2003 at 11:09:32 AM EST

For a long time I've been thinking about creating a community for educational purposes but I always had three problems:

      1) I'm not an expert on any topic.
      2) It's hard to start a community.
      3) I'm a student so I don't have much time.

I taught for a time (not in a school nor a college/university) and I like to teach because I learn too. As you can see, I'm not an English speaker so that's another problem for me. If someday you start a community, I would like to participate. I'm a programmer/computer addict.

Good luck with it
Lot to see, lot do... Yago

You know (3.00 / 1) (#120)
by Control Group on Wed Oct 01, 2003 at 02:09:55 PM EST

As you can see, I'm not an English speaker

Actually, I couldn't have seen that. Like many people I've encountered online for whom English is a second language, you seem to have mastered it better than most of us native English speakers.

***
"Oh, nothing. It just looks like a simple Kung-Fu Swedish Rastafarian Helldemon."
[ Parent ]

Thanks for the cumpliment (3.00 / 1) (#124)
by YagoBlade on Thu Oct 02, 2003 at 09:57:56 PM EST

Sometimes, as many others, I doubt about my English level and I thought that I used some "text book" English. I still have a lot to learn.

Thanks
Lot to see, lot do... Yago
[ Parent ]

Why don't you consider what already works? (3.00 / 1) (#110)
by Smerdy on Tue Sep 30, 2003 at 01:39:35 PM EST

I don't think there's a need for an intentionally organized and new system like this to solve this problem. I think human society has already evolved the essential features of the solution. Within large units of knowledge, it avoids the problems of inconsistent structure that you mention. What I'm referring to is the common usage of textbooks written by single authors or small groups of them.

The problem is that most of them are not freely available, preventing many people from accessing them and forcing those who can get copies to obtain those produced which much more overhead than would be necessary to download them from the Internet. Therefore, I would say that this is a social problem related to intellectual property laws and social attitudes towards intellectual property.

So imagine that all textbooks were placed online, and that we had a reasonably good search engine for finding them. Let's look at how that would fit your criteria.

Topics broken into modular lessons: The vast variety of textbooks today certainly provides many different separately digestible units of information, both in the separate books and in their separate sections and chapters.

Tree like structure organisation of information: Textbooks' forwards often provide suggested orderings of their chapters, defining a tree structure within the books. There's also an implicit requirement graph between textbooks that isn't too taxing to deduce.

Defining assumptions: Textbook forwards generally also enumerate the assumptions made of readers.

Peer review: Textbooks today are created to be bought by students, based on the choices of this teachers. This sort of market economy is a pretty effective kind of peer review, with the teachers as the peers of the authors.

Lesson specific comments: This is pretty easily grafted onto traditional textbooks in online format. A lot of uses of this would probably be replacements for face-to-face interaction with teachers.

Regular lesson reviews: Textbooks often go through many revisions, reflecting changes in their fields.

Strict editorial guidelines and Funding: These fit in with the market economy I mentioned before, one forced by it and the other enabled by it for successful books.

So, in summary, I'd say that the real problem is building a social structure where potential textbook authors are motivated to release their works for free and where they can hope to survive comfortably despite dedicating large chunks of their time in this way.

Talk to the publishers (3.00 / 1) (#126)
by domovoi on Fri Oct 03, 2003 at 09:18:51 AM EST

I'd submit that it's not the authors. Writing texts pays crap. The publishers make beaucoup bucks selling the hell out of a text, and then replacing it or issuing a new edition every two years. Yes, these editions are frequently necessary, but not as often as they're undertaken.

I'd suggest to the OP that, in many ways, wiki approaches what he's wishing for.


------------------------------
This is not my signature line.
[ Parent ]
Connexions at Rice U (3.00 / 1) (#111)
by jhoffa on Tue Sep 30, 2003 at 02:47:47 PM EST

The connexions project uses topical modules that each stand alone, but also can be organized into "courses." All the content is on the Creative Commons licenses.

From the front page:

The Connexions Project is a collaborative, community-driven approach to authoring, teaching, and learning that seeks to provide a cohesive body of high-quality educational content to anyone in the world, for free. The project involves two basic, interrelated components: (1) a Content Commons of collaboratively developed, freely-available material that can be modified for any purpose, and (2) open-source software tools to help students, instructors and authors manage the information assets in the Content Commons. Connexions provides an open, standards-based approach for sharing and advancing knowledge to benefit the global educational community.

Check out the "explore repository" at the bottom of the page to get a feel for the modules.

No mention of assessments (3.00 / 2) (#116)
by willijar on Wed Oct 01, 2003 at 03:21:55 AM EST

In the proposal there is no mention of assessment or feedback on progress which I consider essential for learning (rather than just informing). One technique I use in my online tutorials is to have assessments attached to each concept (single page). The results of these assessments are used to provide guidance through the tutorials using a fairly simple but affective probabilistic network. This could I think be integrated into what you propose. Even better have different prople write the assessments than the provide the information - then a measure of how effective a particular tutorial is can be measured by the average students performance on the assessment. You could perhaps even use the extreme programming idiom - write the tests first and then have people write the tutorials to enable students to meet the measured perfromance criteria.

A second point I would take issue with is the use of a strictly heirarchical approach - not all information can be represented in this way. Additionally you may wish to refer to the same concept/page in several different tutorials. Again I have solved this by treating the tutorials as simply a collection but allowing concepts to specify children - the same concept can then be used in differing heriarchies.

Not necessary for programming (3.00 / 1) (#118)
by Smerdy on Wed Oct 01, 2003 at 11:35:29 AM EST

I never needed stated "assessments" to learn programming. As long as you are learning for the purpose of creating something, you will set your own goals and learn through working towards them.

[ Parent ]
disagree (3.00 / 1) (#121)
by Work on Wed Oct 01, 2003 at 03:35:45 PM EST

if you're going to be using abstract advanced concepts such as discrete math, an assessment is a good thing. Its very easy to get an idea about those you think is right, but can be wildly incorrect.

As CS students the world over know :)

[ Parent ]

disagree (3.00 / 1) (#123)
by Smerdy on Thu Oct 02, 2003 at 03:06:55 AM EST

I said "programming," not "CS." I was just providing an example from my experience showing that externally imposed assessments are not a universal requirement.

[ Parent ]
in programming the assesor is the complier (none / 0) (#133)
by lukme on Wed Oct 22, 2003 at 12:46:30 PM EST

There is no such assesor for math or other subjects which would be necessary for your environment.


-----------------------------------
It's awfully hard to fly with eagles when you're a turkey.
[ Parent ]
Why don't we try it? (3.00 / 1) (#125)
by YagoBlade on Thu Oct 02, 2003 at 10:08:41 PM EST

Something essential for your proposal is people interested. As you could see, you have a lot of people posting on your thread and many approving it. Why don't we try to start a learning community? It would be hard at the beggining, but it's worthy. I don't have much time right now, but if it becomes reality I'll do my best and make some room for it in my schedule.

A lot of people are going to answer this post telling that it's not so easy and things like that, but it's impossible if you don't even try. Don't be ambitious, step by step and something big will come up.

Some interested on starting a community? REALLY INTERESTED?

Lot to see, lot do... Yago

Please go ahead (3.00 / 1) (#128)
by xasmx on Sat Oct 04, 2003 at 04:33:24 PM EST

The easy part is asking others to contribute their continous time to create something and to promise to share the fruit of their labour.

The hard part is contributing lots of your own free time (of which most of us have very little) to make the fruit grow in the first place.

Technically, I'm sure we could create it. Design wise, likely we'd have to try out couple of different things, but could do it. Content wise we'd have to put a continous effort to maintenance, to get a critical mass to adopt it.

Please go ahead. We'll be there when you have created something.


[ Parent ]

perhaps the collaborative media foundation (none / 0) (#137)
by Schnapp23 on Wed Nov 05, 2003 at 01:21:11 PM EST

Maybe CMF can help to get a project like this off the ground.

Anyway, I'd help out, even in the bootstrap process. Is anyone else really interested?

[ Parent ]

Information in isolation (3.00 / 1) (#127)
by DaChesserCat on Sat Oct 04, 2003 at 02:07:47 PM EST

This touches on something which I have been thinking about to a very great deal.

I have three kids in school; one of them starting high school, the others are in middle school. They are getting pretty good at digging through their textbooks and working "recipe-style" at formulating answers, but if I ask them about something they covered a couple days ago, they can't answer. They don't have their textbook in front of them, so they can't "cook up an answer." This bothers me, because it tells me that they aren't really learning anything.

I've been thinking along the lines of creating mini-lessons, each one with specific prerequisites. The idea is that, when they've polished off a certain lesson, they take a quiz on the subject. Passing the quiz gets them a certain amount of points, which can be traded for video game time, TV time, etc. Failure to pass the test gets them nothing.

They can go back and review a lesson, but taking a test they've already passed gets them considerably fewer points. I'm looking at the possibility of some kind of markup language to handle this, as I'd like to be able to run the client on:
  • a web browser
  • a handheld (imagine downloading a few lessons to a Palm, or maybe some kind of module for a GameBoy Advance, then taking a road trip)
  • a video game console (thinking about getting an old DreamCast, getting BSD or Linux running on it, and using that; it would be cheaper than a full-blown computer)
Further lessons would require the student to actually use something they've covered in a previous lesson. It would tie history, math, social studies, languages, etc. together. For example, you could be studying the Romans, and show how Latin has contributed to our use of English, and how roman numerals were superceded by Arabic numbers.

That last point is critical. If you learn how to add, but you never learn any practical use for it, you will quickly forget it. If you learn how history has shaped the language we use, history becomes more interesting. The current schools and classes, with their rigidly defined subjects, are teaching information, separate from everything else. There needs to be a way to tie everything together, because it truly is connected. The sooner people realize that, the sooner they realize that they can learn just about anything from observing or learning about something else. That's when they truly gain knowledge.

Trains stop at train stations Busses stop at bus stations A windows workstation . . .
Idea on markup languages (none / 0) (#131)
by domovoi on Mon Oct 06, 2003 at 11:51:31 PM EST

XML seems like the obvious choice. Some flip-quick looking around rendered this, which might work.
------------------------------
This is not my signature line.
[ Parent ]
That's not the spirit (none / 0) (#130)
by YagoBlade on Mon Oct 06, 2003 at 08:31:25 PM EST

I know I was very optimistic and I think in the previous comment I was clear about it. I'm not telling that everything, I'm just offering me from the beggining and encouraging Stick (the author) to start something. I'm not a good boss for this job, I don't know the "technology" available, so I ask the author, who seems to have more knowledge about this topic, or anyone to define it. I offer myself to maintain and to write articles or even to settle everything up if someone guide me.

Anyone??
Lot to see, lot do... Yago

RFC: mobile network of learning objects (none / 0) (#134)
by day on Fri Oct 24, 2003 at 10:49:06 AM EST

[this is clipped from a letter i wrote]

have you read any of the recent instructional technology work, such as that on "reusable learning objects"? Eg a good and short paper is http://wiley.ed.usu.edu/docs/lo_do.pdf (linked from http://www.reusability.org/blogs/david/archives/000066.html)

the RLO terminology is imprecisely used so i think i'll avoid using it again here: but there are some interesting directions for development in the field.

my concern is that many self-contained lessons, such as tutorials, are of a similar form to games, movies, and other interactives. Currently the ideal medium available appears to be Flash. When categorizing interactive flash animations I don't see a clear boundary between the educational, entertaining, and the social.

so i'm thinking there's a need for a distribution network for self-contained flash clips.

the protocols and classification should be highly structured, so a googlish beast isn't needed to browse and search.

all content should be Creative Commons or similar in license.

eg this could be implemented on the internet, with the navigators (clients) themselves being applications in Flash.

ideally the clients will be mobile, but it should work on desktops too.

ideally the client display would use electronic ink, but i who knows when that'll be viable. And VRD likewise. Until then, something like the old Pogo or the forthcoming Tapwave may serve the purpose.

ideally the client machines would be continuously and wirelessly connected to this network, of course, but meanwhile it should be easy to transfer clips out of desktops using usb pens, flash cards, etc.

the user experience could be: Jim, reclining on a park bench and thinking about his forthcoming french restaurant date, loads the flash navigator application on his handheld and browses the "learning > languages > French > beginner" section. He clicks on a interactive called "French dinner menus", and it downloads while he gets his bluetooth headphones on.

damien
lisbon, portugal
--
babelguides.com << world literature in translation >>

I appreciate the guidelines (none / 0) (#135)
by Persol on Mon Oct 27, 2003 at 06:28:31 PM EST

I am currently working on a site which aims to address the lack of a good and free site for online education. I've realized that the hardest part is the creation and organization of lessons.

For a wide range of material to be taught, it must be open to public input. As we all know though, public input does not equal quality. May current solution is to allow ANYBODY to create a class. However these unreviewed classes will be at the bottom of every class search and listing. When the moderators and instructor agree that a class is substaintially developed, it will be released as 'Official'.

I intend to add the "Lesson specific comments" as the article suggests. I hope to find more useful suggests in the many comments.

---
MangaPug.Com - A community for manga artists.

Proposal for a collaborative learning system | 136 comments (122 topical, 14 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!