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]
Amazon Customers' Recommendations scammed

By tandoor in Internet
Sun May 26, 2002 at 04:13:51 AM EST
Tags: Books (all tags)
Books

This "clever" author of a $3 Self Help PDF has written a program to put his book in as a recommendation 12 times, on every single top seller at Amazon (at least random collection I checked in the first 500 top sellers).

As a result he is now the #3 best seller on Amazon.


Based on the feedback reviews, this has been going on for over a week.

Seems like a complete breakdown of the Amazon customer recommendation system. Yet... no action by Amazon, and it's starting to spread.

For instance, I was looking up the new book by Alton Brown (Good Eats) Food + Heat = Cooking and found 8 recommendations (2 as a replacement!) for yet another self help book: Birdseed Cookies: A Fractured Memoir.

Looking at this best selling How I Play Golf by Tiger Woods, I find the top recommendation (towering over any others with 5 duplicate recommendations) for Open Your Mind, Open Your Life: A Little Book of Eastern Wisdom.

Not exactly a close topic.

Not suprisingly, the same book has multiple recommendations in numerous random best sellers.

Amazon does have a 24hour hold on customer recommendation submissions (I tested it, none of my submissions from about 8 hours ago have appeared yet). But it seems there is no human review, or even simple flood checking going on.

Sponsors

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

Login

Poll
Do you trust Amazon ratings?
o Never, they are mostly fake, and I don't trust the user reviews 19%
o The ratings are junk, but some user reviewers are trustworthy 47%
o I trust the ratings when large numbers of users have participated 18%
o I often use the ratings to help decide what to buy 10%
o I always buy only highly rated items, and always trust the ratings 0%
o Amazon is god, they can do no wrong 4%

Votes: 147
Results | Other Polls

Related Links
o Self Help PDF
o Food + Heat = Cooking
o Birdseed Cookies: A Fractured Memoir
o How I Play Golf
o Open Your Mind, Open Your Life: A Little Book of Eastern Wisdom
o Also by tandoor


Display: Sort:
Amazon Customers' Recommendations scammed | 65 comments (40 topical, 25 editorial, 0 hidden)
Amazon customer recommendations and reviews (4.50 / 8) (#19)
by IHCOYC on Sat May 25, 2002 at 05:15:35 PM EST

I have posted more than a hundred reviews to Amazon.com under the nickname holy_roman_emperor. That the feedback and ratings system is flawed, partisan, and open to abuse is fairly obvious, and not news. If you wish your Amazon mojo to rise, you will do well not to say anything mildly negative about any product that might have vocal fans. Still, this does not strike me as new or particularly surprising.
--
"Complecti antecessores tuos in spelæis stygiis Tartari appara," eructavit miles primus.
"Vix dum basiavisti vicarium velocem Mortis," rediit Grignr.
--- Livy
Flawed and partisan (3.25 / 8) (#20)
by dieMSdie on Sat May 25, 2002 at 05:35:25 PM EST

That the feedback and ratings system is flawed, partisan, and open to abuse is fairly obvious, and not news.
Oh, Amazon's system! For a minute there I thought you were talking about /. and K5! ;)

[ Parent ]
really? I never have problems. (none / 0) (#56)
by Shren on Tue May 28, 2002 at 09:56:08 AM EST

I don't see myself as part of the K5 mainstream, in terms of either popularity or ideology, yet I still find that the good things I post get rated up, and the occasional bad thing gets rated down. Can you suggest any specific victims of the K5 cabal?

[ Parent ]

Customers' Recommendations is much more abusable (4.50 / 2) (#26)
by tandoor on Sat May 25, 2002 at 10:57:15 PM EST

I agree, false reviews are nothing new at Amazon, at least they have some sort of user moderation system that allows you to rate the quality of a review.

But this Customers' Recommendations feature has no such moderation, and apparently no human checks either. And it's a much more powerful feature than user reviews. It appears before user reviews as part of the recommendations. When I first saw it, I assumed it was some kind of derived suggestion much like the the other recommendations system. I only looked into what it actually is when I saw such bizarre recommendations.

What is particularly powerful about Customers' Recommendations is that it allows users to pull in customers from other books. You can link into your book from an apparently unlimited number of other books, with no moderation of any kind, either by Amazon, or by users.

[ Parent ]

Is that really important? (4.57 / 7) (#21)
by mami on Sat May 25, 2002 at 06:12:27 PM EST

Are the customer reviews on Amazon really a determining factor in your book buying decisions?

I could understand, if you were inspired to buy a book by reading about it in WP, NYT or on C-SPAN or any of your fan subject web sites or journals. But reviews on Amazon.com?

May be I would read them to look if they reaffirm my own expectations about what I think the book might be all about. I rather find out if other people thought negatively about the book than listening to their praise.

If I intend to buy a book, I usually have already some positive reasons to do so. They are not triggered by some reviews on Amazon. Unless you are not a book author and want to sell your book, I don't understand why you would care.

Is it really all that important? (apparently is) (4.75 / 4) (#27)
by tandoor on Sat May 25, 2002 at 11:09:08 PM EST

Apparently it is important to many Amazon customers, since the Amazon sales rank for that PDF skyrocketed to sales rank #3 after posting the Customers' Recommendations (it's #5 right now).

Pretty impressive results for something that everyone ignores!

and yes, I do tend to notice reviews and rankings at Amazon, if only as a starting point before more research into a book. At one time, I might even buy a book if there were enough positive reviews and it sounded interesting (but with shelves accumulating the to be read, I try to avoid that now).

and yes, I do use the previous recommendation system, if only on lark to see what the heck Amazon thinks I would like based on the random collection of books I've rated or bought. Based on my experiences with "MovieCritic"'s (now defunct) and "NetFlix"'s recommendation systems, I don't expect much.

[ Parent ]

Of course it is important (4.00 / 3) (#38)
by mami on Sun May 26, 2002 at 09:07:32 AM EST

Sorry, I admit, I made my comment before reading your article carefully. It's important and I appreciate you pointing out their system's flaw.

I was ignorant, I seldom use reviews before I buy books. I read them in the bookstore and usually know after two chapters, if I want to take them with me or not.

[ Parent ]

Google versus Amazon (4.50 / 6) (#24)
by danny on Sat May 25, 2002 at 08:40:00 PM EST

There was a time when I worried that Amazon would get a stranglehold on book information - its reviews would make it everyone's first port of call for information about books, and enable it to collect still more reviews, etc. to maintain that position.

But two things prevented that. The first was so many newspapers, etc. going online with their book review archives and the growth of Internet-only reviewing (to which I've made my own contribution) and the development of Google as a way of searching through that easily.

The second was the out of control growth of content on Amazon - even without the quality control problems this article raises, who needs, or can possibly use, 500 customer reviews on some titles?

Danny.
[900 book reviews and other stuff]

What is uniq is the "Pull" factor, like (4.50 / 2) (#28)
by tandoor on Sun May 26, 2002 at 12:24:11 AM EST

Consider each "Customers' Recommendation" an ad, costing nothing, which you can place under every top selling book you want. It costs you nothing but the time (or effort if you write a bot), and gains you wide exposure. There is already a "war" of sorts developing for the top spot. Looking at a couple top sellers, I see the fake recommendations up over 30 (yesterday I didn't see many over 12, a few 20s).

These are not really "reviews", but are supposed to be "if you liked this book, then customers are Amazon think you might like these books". It's a nice idea, but dates back to a different kind of Amazon (if it ever existed). And there really are some real links in there made by real fans of a particular genre, usually with a rating of 1, or 2, or 3 (I saw one legit 5). Sad to seem them down at the bottom.


[ Parent ]

based on actual sales? (3.00 / 1) (#32)
by danny on Sun May 26, 2002 at 08:06:47 AM EST

I thought Amazon's recommendations were based on actual sales - "people who bought this book also bought X, Y, and Z". That's not impossible to spoof, of course (cf the author who bought 30 000 copies of his own book for a spot on the bestseller lists), but it's a lot harder.

Danny.
[900 book reviews and other stuff]
[ Parent ]

Now two kinds of recommendations (4.00 / 3) (#33)
by tandoor on Sun May 26, 2002 at 08:14:52 AM EST

Yes the original recommendation system cannot be spoofed.

It's the new "Customers' Recommendations" system that is so wide open to abuse. (seems to have no limits or protections of any kind)

[ Parent ]

Really? (3.00 / 1) (#51)
by ti dave on Mon May 27, 2002 at 12:14:43 PM EST

(cf the author who bought 30 000 copies of his own book for a spot on the bestseller lists)

Who did this?
That's an interesting gamble!

"If you dial," Iran said, eyes open and watching, "for greater venom, then I'll dial the same."

[ Parent ]
Not Really a Gamble (4.25 / 4) (#52)
by edibiase on Mon May 27, 2002 at 12:41:28 PM EST

If I recall correctly, he just returned all of the books he bought right after he bought them. The purchasing of the books was counted toward making the work a bestseller, and the returns were never factored in.

Apparently this is standard industry practice.

[ Parent ]

How would you fix the Amazon Customer Recommends? (4.80 / 5) (#29)
by tandoor on Sun May 26, 2002 at 12:38:57 AM EST

Here is how it currently works:

You enter the "ASIN" (Amazon stock number) into a single field, and click a button. You need to be logged in as an Amazon user. No other data is required, and you can enter multiple recommendations for different items.

There is no apparent limit on how many different books you add a recommendation to.

In 24 hours, the site is updated to reflect your additions.

Obvious changes would include:

Track number of recommendations made to a single book (both in one day, and over time). You need humans to look at the numbers and respond. It won't stop those who just want to undermine the system with random recommendations, but it will stop those using the system to promote specific books.

Limit how many recommendations can be made from a single IP address over a period of time. Say 20 per IP per day. But then what about big companies where employees are behind a firewall, do you shut them out?

Ok, then limit how many per account per day. But there is no limit on how many accounts you can create from one IP, all you need is an email address (easy).  So you use multiple accounts. (likely already going on, certainly does for reviews).

So you limit Customer Recommendations to only users who have held accounts at Amazon for 6 months, and have made a purchase. Seems a bit severe, but how else do you control the multi-account problem?

Add moderation to recommendations, rank lower the recommendations where majority disagree with the recommendation.

You could aslo force (2.33 / 3) (#37)
by mami on Sun May 26, 2002 at 09:00:29 AM EST

the people who want submit revies to write at least 3000 words reviews or something like that.

Usually that is something only very determined people would do for free, and if it would be a fake attempt of propaganda for or against a book, it's easier to detect in a longer review than in a short two to three paragraph piece.

[ Parent ]

3000 word reviews? (4.33 / 3) (#39)
by tandoor on Sun May 26, 2002 at 09:38:22 AM EST

I doubt half of 1% of the reviews on Amazon are anywhere near that long.

Besides, there are clever work arounds, review generators, boilerplate reviews, and quoting existing positive reviews from other sources. (not to mention the powerful incentive to the author to create even 2 or 3 positive reviews for a smalltime book).

[ Parent ]

well then, don't read the crap :-) (1.00 / 1) (#40)
by mami on Sun May 26, 2002 at 09:52:20 AM EST

I doubt half of 1% of the reviews on Amazon are anywhere near that long Exactly.

[ Parent ]
How about (4.00 / 2) (#49)
by jmzero on Mon May 27, 2002 at 10:12:08 AM EST

Only users who have bought something can do reviews?  

And a single abuse (which can be reported by anybody and is checked by Amazon staff) means they cannot review or recommend books for six months.

Amazon could solve this problem if they wanted to.  And they would want to if these were negative (sales undermining) reviews instead of positive ones.
.
"Let's not stir that bag of worms." - my lovely wife
[ Parent ]

how you can use the ratings (2.00 / 1) (#30)
by nodsmasher on Sun May 26, 2002 at 03:46:14 AM EST

if there is some concensus in how somthing is rated, ie if every one who gives it a 1 says "(insert authur here) sucks and he will write nothing good" but all the people who give it a 5 say "its a masterpiese of imperfect perfection" then you might think be able to use that to your advatage
(example used)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Most people don't realise just how funny cannibalism can actually be.
-Tatarigami
How to fix recommendations... (4.00 / 1) (#31)
by leviramsey on Sun May 26, 2002 at 07:02:02 AM EST

It is immediately apparent that this system has a massive potential for abuse. Such abuse is occuring. I feel that there is a partial solution to this problem (as it's partly a social problem, technology cannot solve that portion...).

We can view recommendations as a weighted undirected graph which allows multiple distinct edges between vertices. Each edge corresponds to a recommendation by a customer/user.

Each customer/user has a recommendation power rating which is one plus the number of items purchased over the past six months. The weight of each edge in the graph is equal to this power rating divided by the number of recommendations. Because power ratings only cover purchaes in the previous six months, it is logical for the recommendations to expire after this time.

The strength of a recommendation from A to B is the sum of the weightings of all edges from A to B.



Power of ratings based on purchases? (4.00 / 1) (#35)
by tandoor on Sun May 26, 2002 at 08:31:12 AM EST

My problem with this is:

Will the people who spend the most (in dollars) be the ones who's recommendations have the most value?  I don't think so. For instance I wouldn't be suprised if some of the top reviewers buy most of their books at places like half.com, or bookfinder.com, or the local Costco.

If you are going to go for a capital based system where you have a certain worth you can "spend" on recommendations, I would base that worth on:

1. How long you have been a customer
2. How many reviews you have written
3. How users have ranked your reviews (skewed twords those with more worth)

But this comes back to, why restrict honest users from making many recommendations? If someone is an expert on cookbooks, wouldn't it be good to have them posting recommendations all over the place? and wouldn't you want those recommendations to have a fairly high "value"?

Besides there is still the serious problem of multiple accounts used for create fake reviews, and to rate all those reviews highly, and to mark down any negative reviews (for instance marking down any reviews that point out the current scam).

Amazon needs to start running some serious analysis of user activity. When 12 users all post the same recommendation for hundreds of books, wouldn't you think that would stick out in any kind of reasonable activity analysis program?  Or when those same 12 users suddenly all rate every negative review as unhelpful, each time a new negative review appears?

[ Parent ]

would be awful (2.25 / 4) (#36)
by mami on Sun May 26, 2002 at 08:54:23 AM EST

Will the people who spend the most (in dollars) be the ones who's recommendations have the most value?

If that is so, that's despicable.

[ Parent ]

Why, exactly? (3.33 / 3) (#41)
by Giant Space Hamster on Sun May 26, 2002 at 01:06:25 PM EST

I don't think it will be that bad. Look at it as "putting your money where your mouth is".

I think someone who spends a large amount on books might be more trustworthy (in a Amazon recommendation sense only) than someone who's never bought anything, and may be only making comments from the peanut gallery.

Of course, the guy who's never bought books may be an expert, but it's really hard to judge that.

-------------------------------------------
The whole problem with the world is that fools and fanatics are always so certain of themselves, but wiser people so full of doubts.
-- Bertrand Russell
[ Parent ]

What ? (3.50 / 4) (#43)
by mami on Sun May 26, 2002 at 02:10:57 PM EST

I know a lot of people, who can't afford to buy many books, but read tons of them through libraries.

If someone is an expert or not, you are able to make a judgement about, if he had to write a 3000 word book review.

So, to make it dependable on how many books one has bought from Amazon is IMHO a cheap step to encourage sales, one I don't particularly like.

[ Parent ]

Yes, but (4.00 / 2) (#44)
by Giant Space Hamster on Sun May 26, 2002 at 03:26:06 PM EST

See the point I'm trying to make is that you can't really tell someone who reads a lot (but doesn't buy) from someone who's just spamming.

But if they valued the book enought to buy it (which Amazon can confirm), that lends them a little bit more credibility.

Talk is cheap, after all.

-------------------------------------------
The whole problem with the world is that fools and fanatics are always so certain of themselves, but wiser people so full of doubts.
-- Bertrand Russell
[ Parent ]

Talk shouldn't be cheap (2.66 / 3) (#45)
by mami on Sun May 26, 2002 at 04:43:05 PM EST

unless you allow it to be. It's Amazon, who wants to increase customer service, increase sales and provide incentives to look up their site, right?

So, instead of allowing spamming of reviews to take place and messing up the meaning of the review system, they should go the old fashion way and make a fair cost analysis about how much they can spend on book reviews by serious, independent reviewers. If they can't spend the money on it, then it's better to have less or no talk than cheap and biased talk.

I would neither rate the professional reviews, nor would I rate reader's comments. Let them speak for themselves.

May be I overlook some aspects here, I don't know.

[ Parent ]

Amazon is all about biased talk (5.00 / 2) (#62)
by Detritus on Fri May 31, 2002 at 02:34:10 PM EST

If they can't spend the money on it, then it's better to have less or no talk than cheap and biased talk.

Biased talk is what the Amazon review system is all about. For example, when I read the reviews for Chomsky's 9-11, I want to hear what an Arab man who thinks that all Americans are pigs has to say, and I also want to hear what the typical Christian "bin Laden is going to Hell" person has to say.

For me, it is much easier to sort through two articles tilted toward two different sides and find a middle ground than to read a "serious, independent" article devoid of excitement and genuine interest in the subject.

Kings and lords come and go and leave nothing but statues in a desert, while a couple of young men tinkering in a workshop change the way the world works — Havelock Vetinari
[ Parent ]

Re: Power of ratings based on purchases? (3.00 / 2) (#47)
by leviramsey on Mon May 27, 2002 at 03:52:19 AM EST

Under my plan, only the quantity of books purchased (whether new or used) is considered. Someone who buys 30 books totalling $100 will count for more than someone who buys 10 brand new hardcovers totalling $290.



[ Parent ]
A better answer... (4.00 / 1) (#46)
by MickLinux on Mon May 27, 2002 at 02:56:17 AM EST

A better answer, perhaps (not my idea) is to allow users to rate books both on an absolute scale, and on a scale "if you liked A, you'll also like B." Then, for a given user you can give him a trust level. You get recommendations from your own trusted reviewers, and 80% of his trusted users, and 64% of their trusted users, etc. That is a bit easier to do than it seems, because your own rating incorporates the fractional ratings of everyone else. Ideally, you can also see who others trust. As this happens, everyone will get a bunch of recommendations that they actually believe.

I make a call to grace, for the alternative is more broken than you can imagine.
[ Parent ]

i've never used their recommendations (3.50 / 2) (#42)
by infinitera on Sun May 26, 2002 at 02:08:57 PM EST

User created lists are my real resource there. Generally, for every book I need for a project, browsing through others' lists gets me 10 more good ones. Distributed processing, yay! ;)

Libraries do this so much better (4.00 / 2) (#48)
by blisspix on Mon May 27, 2002 at 04:02:24 AM EST

You know, if Amazon had used some long established tools used by libraries, their system would work a whole lot better. As many of you would know, libraries help you to find things on their catalogues by using subject headers, a very very long and complex list of standardised terms used to describe an item. They cannot be tampered with or modified, and many books include this data in their publication details at the front of the book. For an example see here - http://www.lib.duke.edu/libguide/fi_books_sh.htm I think that they have been underutilised in commerical environments, especially for the sort of thing that Amazon is doing. Using LCSH would also allow libraries to create a list of subjects that they want to get updated book lists for. Say, when a new book on a subject is published and available for purchase, they could use LCSH via z39.50 on amazon.com and bingo! purchased. This would also help simplify book purchases especially in large libraries.

Old friend, you know who you are! (1.00 / 1) (#50)
by GoStone on Mon May 27, 2002 at 11:35:39 AM EST

I looked up an old friend recently and found an Amazon page showing he had written five books. They all had glowing reviews. I'd love to know if they are real, but I suspect he wrote them himself, the sneaky egoistic bastard. Good luck to him I guess.


Cut first, ask questions later
Corrections (5.00 / 2) (#57)
by Waldo on Tue May 28, 2002 at 04:36:12 PM EST

I need to correct a few problems in this article.

1. Birdseed Cookies is my mother's book and it is most certainly not a self-help book. It's exactly what it's called: a memoir. You did bother to read the title before writing this, didn't you?
2. Searching for Answers is not "the #3 best seller on Amazon." It's the #3 best-selling PDF. Given that PDFs sell for shit, that's pretty much a meaningless number.
3. This recommendations system has only been in place for a month.
4. The self-publishing community has been very active in recommending the books of their friends. In the early days of this system, it's easy for a very small number of people (as you can see) to affect ratings strongly. In my mother's case, she's on a few discussion lists for e-publishers, and I know that they all tend to vote for one another on Amazon. Further, my mother was in the New York Times last week, and is a regular featured essayist on National Public Radio. Her response to all of her fan mail in the last month has encouraged people to vote for her book on Amazon.

The reason that there's no "flood checking" is because, at least in my mother's case, there's no flooding. I'd know -- if she were using a program to flood, I would have been the one to write it. :)

tandoor, this is a democratic system. It's just like elections: if 12 people vote, and you're not one of them, then you don't get to gripe. With a system this new, I guess you could say that a book is "towering over any others with 5 duplicate recommendations," but how about you give it 60-90 days until people discover and start using this system?

I'm only disappointed that I didn't spot this story in the queue and make these corrections before this ran.

-Waldo Jaquith

Others do the dirty work, then it's ok? (4.00 / 1) (#58)
by tandoor on Wed May 29, 2002 at 01:43:05 AM EST

1. Birdseed Cookies is my mother's book and it is most certainly not a self-help book. It's exactly what it's called: a memoir. You did bother to read the title before writing this, didn't you?

The reviews compare it to the Chicken Soup type books which are considered "self help". I'm sorry if you find that an insult.


2. Searching for Answers is not "the #3 best seller on Amazon." It's the #3 best-selling PDF. Given that PDFs sell for shit, that's pretty much a meaningless number.

The Amazon page says "Amazon.com Sales Rank: 4" (right now). That's what I refered to. Did I say #3 best selling BOOK on amazon? (no, I thought not).


3. This recommendations system has only been in place for a month.

That's longer than I thought. In my messages I refer to it the abuse dating back around a week or so based on suppressed reviews.


4. The self-publishing community has been very active in recommending the books of their friends. In the early days of this system, it's easy for a very small number of people (as you can see) to affect ratings strongly. In my mother's case, she's on a few discussion lists for e-publishers, and I know that they all tend to vote for one another on Amazon. Further, my mother was in the New York Times last week, and is a regular featured essayist on National Public Radio. Her response to all of her fan mail in the last month has encouraged people to vote for her book on Amazon.

I hope she hasn't been encouraging them to vote for her book by creating referals from completely unrelated books. Because that is what is going on. For instance under a "New Kind of Science" by Wolfram, their are 11 "Buy this instead" recommendations for your mother's book. Recommendations for your mother's book also appear in all the best sellers I checked. But she is already loosing in most cases because others have stepped up their voting efforts. For instance "Breaking the Pattern: The 5 Principles You Need to Remodel Your Life" now has 40 "instead of" recommendations under "A New Kind of Science".


The reason that there's no "flood checking" is because, at least in my mother's case, there's no flooding. I'd know -- if she were using a program to flood, I would have been the one to write it. :)

There absolutely is flooding going on. Just because your mother is not doing the flooding, but is asking others to do it for her does not mean she is not abusing the system.


tandoor, this is a democratic system. It's just like elections: if 12 people vote, and you're not one of them, then you don't get to gripe.

One vote one person? More like 1 account, infinite number of votes.


With a system this new, I guess you could say that a book is "towering over any others with 5 duplicate recommendations," but how about you give it 60-90 days until people discover and start using this system?

So all authors need to round up groups of Amazon users to put their recommendation in for every book in the top 2000 (or 5000) sellers? And the "winner" with the most fans will have their recommendations put at the top? I suspect every recommendation will soon point to this book (once word gets out). Or maybe PR companies will sell "recommendation voting" services to get your book listed as a recommendation on a few best sellers.


I'm only disappointed that I didn't spot this story in the queue and make these corrections before this ran.

It sounds more like a defense/justification than "corrections". "Her book isn't self help, she's a real author, so this is ok", "She didn't post those false recommendations, her fans did", "The system will correct it's self once enough people vote (flood) the system with their own recommendations".




[ Parent ]
You Misunderstand, I Suppose (none / 0) (#61)
by Waldo on Wed May 29, 2002 at 09:57:02 AM EST

I hope she hasn't been encouraging them to vote for her book by creating referals from completely unrelated books.

Who says they're unrelated? Have you read them? And what if they were? As a voting based system, if they're unrelated then before long those votes will be overwhelmed and the recommendation will fall away.

Just because your mother is not doing the flooding, but is asking others to do it for her does not mean she is not abusing the system.

Actually, it it does. Unless, by "flooding," you mean "getting referrals fairly."

I ran for office in February. I got a whole lot of people to vote for me by asking them to do so. Was I abusing the system?
One vote one person? More like 1 account, infinite number of votes.

There are many ways for voting to function. That is, apparently, how Amazon has chosen for it to function. (Though, to be fair, one individual can only make a single recommendation from each book, IIRC.)

"The system will correct it's self once enough people vote (flood) the system with their own recommendations".

I see the problem here: you think that a vote is inherently a flood. Anytime that anybody recommends a book, that's a flood. If somebody recommends a book in relation to a few others, than whoa nelly, things are getting out of control.

The fact that Amazon allows a single user to recommend a single book in relation to 2-3 others indicates that this is an acceptable use of the system to them. If it weren't, they wouldn't allow it, would they? People are allowed to recommend books in relation to others. Some people have recommended Birdseed Cookies in relation to others. You're yet to establish how this is in any way an abuse of the system.

Perhaps more relevantly, it's just a recommendation system on a company's website. It's just out of beta. Nobody's using the thing yet. You're getting all bent out of shape over nothing, I promise. Give it 6 months and this will be a sufficiently highly-used system that it will be nearly impossible for a half-dozen people to count as "a flood" of votes. And, just like now, it will hardly warrant an article on k5.

-Waldo Jaquith

[ Parent ]
Hyperlink error? (none / 0) (#59)
by Rasman on Wed May 29, 2002 at 03:44:32 AM EST

Why do the hyperlinks for How I Play Golf and Open Your Mind, Open Your Life: A Little Book of Eastern Wisdom link to the same URL??

Sorry, I wrote to the editors, they never fixed it (5.00 / 1) (#60)
by tandoor on Wed May 29, 2002 at 05:51:54 AM EST

How I Play Golf by Tiger Woods

Here is the message I sent, no response:

In the story at:

http://www.kuro5hin.org/displaystory/2002/5/25/141522/582

There is URL that should point to "How I Play Golf" by Tiger Woods, but accidently, it was a duplicate of the next URL on the page.

Could you please correct the link to "How I Play Golf" to be:

http://www.amazon.com/o/tg/stores/detail/-/books/0446529311/advice/ref=cm_custre c_gl_more2/002-6500787-2637618

Thank you.


[ Parent ]
Sheep. (none / 0) (#63)
by /dev/trash on Sat Jun 01, 2002 at 05:40:27 PM EST

Do people actually buy things based on what others say?

---
Updated 02/20/2004
New Site
Amazon gave up - story over (for now) (5.00 / 2) (#64)
by tandoor on Sat Jun 08, 2002 at 12:48:44 PM EST

Quote from Amazon:

Many thanks to all of our customers who participated in the What's Your Advice? feature by providing thousands of valuable product recommendations. Your advice helps to improve the personalized shopping experience at Amazon.com. Unfortunately, we witnessed some cases of commercial abuse, which violates our usage policy. We are modifying the feature to put controls in place that will make the What's Your Advice? experience better for you, the customer. Expect to see it back on the Web site soon.

heh, not the first time (5.00 / 1) (#65)
by anon868 on Tue Jul 09, 2002 at 11:51:12 AM EST

Remember, way back in the day, amazon used to have a link that the author could click on & write something about their own book?('I'm the author' or something like that). Anyone could actually click on that link & say they were the author of the book, and write their own review 'by the author', and there were no human checks. It made for some humorous book reviews.
Open a window. No, not that one! One made from actual glass, set in an acual wall, you dork.
Amazon Customers' Recommendations scammed | 65 comments (40 topical, 25 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!