It's the continuation of privacy questions: when you know the information came from a specific person/entity, how do you know to trust it? Obviously if it's on the internet, you don't inherently trust it much more than a talking can of tomato paste. The direct query is, specifically, how do you find out the originator's biases and entanglements?

Question on Objectivity sums up my thoughts in a few mis-spelled paragraphs. Read it.

1. There will be biases on all sides.
2. There is no such thing as objectivity unless robots are doing the review.
3. The Internet is more subjective due to being able to link directly to source material.

Now, blast off into my subjective (and quite probably wrong-headed) view of opinionation... IN THE FUTURE!

First, I'll knock down some of the social ramifications of the thing. Next, a visit to some of the major players involved in opinions as a business. Then, we will whip out the opiniometer and try a real world example of the systems in place. Finally (so as not to scare you right away), the technical side of this whole stew.

Let us start at the Sociology Internet Resources maintained by Western Connecticut State University. The same author also has some handy linklists available on other topics, like economics.

Off-topic thread: I adventured on over to the Essential Principles of Economics hypermedia text and found an understandable page on Giffen goods. Clarified with preference theory graphs, it's quite easy to comprehend lower prices causing less demand in the end.

At last, I struck the mother lode of info-econ. Buried within Berkeley's SIMS site is The Information Economy, an unending source of information on... the economics of the Internet! Go figure. Unfortunately, nothing directly pertaining to opinions and their bias as affected by the digital medium. Plenty of closely related issues covered, though.

Having a difficult time digging up appropriate documentation on bias, without resorting to copying from the books I stumbled through: Bias in Human Reasoning: Causes and Consequences and Bias in Quantifying Judgements.

Here's a chapter on peer review and bias as they play into academic papers from Suppression of Dissent , a resource for keeping arguments and opinions open. Also interesting: Democracy without Elections, a paper on participatory democracy. (More related to collaborative opinion sharing than one might think at first glance.)

Disjointed humour interjection: Are you URO-Curious? You know I am!

Jewish News of Greater Pheonix stays firmly on the ground and basically says, 'Don't fear the reaper, everybody.' Quote:

"It is imperative that new Internet users understand that most Web publishers print generally accurate and reliable information. It is in their interest to do so, in the expectation that visitors will bookmark the site and visit frequently for updates and new content. But fringe political groups and scam artists are placing fraudulent sites and fraudulent facts on their sites. Their interest is fraud, period."

Leonard Grossman raises some highly fortuitous points in his article: Trust or Consequences. Although he's exploring the informational bias side of things, don't ignore that side. It's bigger and much scarier than this consumer goods coverage.

What exactly are the responsibilities of an online writer? How do I make you [as readers] aware of my biases and prior prejudices?

World Opinion is market research. Actually, some of these marketing folks have the right idea; it's all about personal service and getting down in the details.

"Tell them who you are and how you arrived at what you are doing. Show them your picture so that they know it is a real human being behind your operation. It is not about HOW you look. It is about creating an impression of closeness and connection..."

Build your own trust. [The article could be linked to weblogs in extended metaphor; exercise left to the reader.]

Go read this most excellent article on collaborative filtering and "infomediaries" by New Yorker contributor Malcolm Gladwell; you won't regret it.

As Gladwell mentions, there are some issues resolved with marketing and these methods.

"Marketers now play an elaborate game of stereotyping. They create fixed sets of groups--middle-class-suburban, young-urban-professional, inner-city- working-class, rural-religious, and so on--and then find out enough about us to fit us into one of those groups. The collaborative-filtering process, on the other hand, starts with who we are, then derives our cultural "neighborhood" from those facts. And these groups aren't permanent. They change as we change."

"In order to know what someone wants, what you really need to know is what they've wanted."

In the article, Gladwell mentions MovieLens which is based on GroupLens technology. Try it out and see if it doesn't make a believer out of you after you rate 20 movies or so; I am converted and can't wait until systems like this are functional for other types of products.

Introduction to Media Theory: "How do we use communications media, and what do we learn from our use of them?"

This class site has lectures online that cover all sorts of subjects, from individual differences, purposes, and needs to context and selectivity.

Off-topic don't miss: Technological determinism links

Grossman: Guerrilla Journalism
It's the Internet. Lines will be blurred. You will be offended, shocked, bored, and frustrated -- all at the same time!

Consent, American Style: Mass media manipulation of opinions? Never. I don't believe it.

Information Does not Equal Knowledge: Theorizing the Political Economy of Virtuality

"It should be no surprise that confusion about rationality is heightened by the virtual, rather than resolved by it. In this context the conflation of information and knowledge has taken place under the guise of hyperreality and associated practices which promise an interactive immediacy that previous communication technologies could not offer."

Before we drift off into completely relative space and lose the invisible thread, here's one more blast of commentary-free socsci links before delving into the reality of opinions on the web.

Max Weber's View Of Objectivity In Social Science
MIT Media Lab
Resource Center for Cybercultural Studies
UCLA Center for the Study of Online Community
Internet Fraud Watch at the National Fraud Information Center
SocioWeb
Society for Judgement and Decision Making's linklist
Al Roth's Game Theory and Experimental Economics Page
Yale Law opinion linklist Collaborative Filtering at SIMS Community Based Navigation and Collaborative Filtering Systems

Yahoo's consumer opinion listings seem full of whining, and very little constructive resources. Here are some sites trying to make an actual business out of community opinion.

AMAZON

Amazon is more of a last resort spot for finding books than an opinion site.

Jakob Nielsen on Reputation Managers "Amazon.com pioneered the idea of customer reviews, but has been plagued by unreliable reviews (an author's enemies post a flood of negative reviews; followed by the author's friends who post glowing reviews). Also, users never know whether they can trust reviews that are posted as part of a site that profits from selling the product.

Google and eBay avoid these problems by aggregating ratings across a very large sample."

General confusion bleeds through the Amazon comments. It's impossible to mark a trusted source, even with their new-fangled meta-rating system.

The problem? The larger number of folks who have something and review it, the more likely an average consumer will buy it. The idea is: just go nab it, since it's worked for so many friends. I am more curious about relating opinions on obscure things, and that's where minor sampling of a large range falls even flatter than it does in average circumstances.

CONSUMER REPORTS

While this venerable source has no ads, it does come with a monthly cost. Therefore, I can't link you directly to much information. You know this magaizine. You know how it works. You know it's not completely bias-free. That's why there's the Consumer Distorts watchdog.

PRODUCTREVIEWNET

This site is really just a searchable aggregation of external review abstracts. Meaning, it offers no new content; but brings a bunch of information together in a possibly useful way.

REVIEWSEEK

What Product Review Net does for consumer goods, Reviewseek does for computer hardware and software. Great resource if you enjoy drooling over the latest advancements in money-spending.

BIZRATE

Although it smacks of a shopping portal, Bizrate deep down inside is really a polling station that maintains records on major ecommerce vendors. Wondering if ordering that Smezamazog from Wzywaza.com is a safe bet? Go here to find out what other folks who filled out surveys thought.

CONSUMER REVIEW

While some sites strive for that single glowing visage, this company chooses to divide their assets into communities. Helpful if the product you seek falls in one of the categories; they don't do books and music, but cover plenty of things I have never worred about. Baby gear review? Fly fishing review? Sure.

RATEITALL

Basically a format for compiling subjective ratings; at first glance, one might think it's a collaborative net. However, the only way to "relate" to other user's rankings isn't a network at all. Quote:

"If you go to our Current Rock & Pop CDs page, you can find a list of the highest-rated CDs. And, if you want, you can filter the list so it reflects only the ratings left by people your own age and gender!"

The actual interface says "Find out what people like you think". Please, do tell what I think! I'm different, the same as everyone else. All of you are different just like me. Don't rely on age-based segregation! Tear down the walls... ugh, never mind.

PRODUCTOPIA

This site attempts to be everything to everyone, with content running the gamut from top picks, buying tips, consumer guides, to user reviews and discussions. Meanwhile, loud ads and subtle promotions combine with the just plain bad search engine to drive the casual viewer off. I'll expand on why I don't like this site in the experiment.

EPINIONS

Oddly, a google search on epinions returns only the "realname" with no hits; epinion singular works just fine.

Epinions has been beaten to death in the log circuit, so I won't spend too many precious bytes blowing air at it until putting it to use below.

More blabbering than any competitor! On itself. Downer 1. Downer 2. Helping you buy infant death syndrome since 1999 [ev]

Why are some reviews trusted and some not? See for yourself; it is the web of trust in action.

EXP [scribble]

Exp is an interesting site; rather like an Ebay for information. Submit your question or problem to a certain category, and "experts" step up to the block and do their jig. Then you fork over the requisite dough. An open meat market for mini-consultants, if you will.

The most interesting facet that would probably redeem this site if switched? You can read the questions, and see which have been answered, but you can't read the answers. Bring the pocketbook, dear.

There are all the usual problems to the approach, not to mention the simple fact that an obviously money-based information economy goes against the very fiber of the Web's initial intention. Blah. It's good that people are asking some silly questions, though; for example, what's the beat generation? As you will notice, two kind benefactors offered to give a helping history hand to this poor soul. Belly up for the answer.

No lab coat required. You can follow along by picking something you would like to buy someday. (Aww, don't kid yourself, you know you need that Smezamazog.) For the sake of science, I have entered the market for:

a. A hobbyist's digital camera (PDR-M5?)
b. Albert Einstein - The Collected Papers of Albert Einstein (probably the Swiss Years as recommended by AmSci)
c. The new Storm and Stress album (prerelease), or failing that, the highest quality John Zorn release according to the site. If that falls through too, I'll back up and see how their Coltrane resources fare.

Skipping sites that are obviously irrelevant to my impulse buying, off we go.

AMAZON part a part b part c

a. I'm especially interested in the Toshiba PDR-M5 after giving one a test drive, but don't know enough about it to be able to tell what might be wrong with it. Amazon doesn't even know about it, period - their selection is weak. However, the cameras have excellent links to other reviews on sites, and the comments are surprisingly (to me) semi-useful.

b. The "related auctions" for something about "The Swiss Without Halos" and a Swiss Army knife are frightening, there's only one editorial review, and no customer comments. If you click "Hardcover" in the other editions section (this is already the hardcover), you go to Volume 1 - a completely different set of papers. Guess I'd better know what I am buying...

c. This is Amazon we're talking about! Of course they know about the album, and they even have sound clips. Prerelease ordering is easy. No links to reviews, though.

PRODUCTREVIEWNET

a. No.
b. Nope.
c. No. Flunk!

CONSUMER REVIEW part a

a. Also outdated, it doesn't have the PDR-M5 or the Nikon Coolpix 950. While it says that there are 33 reviews for the Coolpix 900, to the untrained eye there only appear to be three. until you click at the very bottom and reach the entire listing.
While there is a rudimentary comment-rating system, it doesn't appear to actually be in much use, as only 2 out of 33 comments are rated. In all though, not a bad resource.

No book or music reviews.

RATEITALL part a part c

a. Do these seem numbers seem skewed or is that just me? Not very many reviewers; maybe that could explain it. The Toshiba PDR series gets a 2.5 rating, but there is no explanation. The site just tracks numbers. Scary.

b. Has books, but not very many.

c. No SnS, no Zorn, but there are a few Coltrane ratings. A Love Supreme gets a 4, with some folks rating it 1. You can't tell who, though! There is one comment on the sad state of affairs:

"What makes me chuckle is that Leann Rimes gets 5-stars for her latest piece of (work), while Coltrane scores a 3.95 for his seminal masterpiece. Curious. This disc moves me like no other."

Obviously, I have the same bias as that writer, because I chose to pull that quote. What other obvious exhibits have I shown thus far?

PRODUCTOPIA part a

a. They have pretty buying guides, but their opinions are severely limited, as you can see by clicking on Toshiba Digital Cameras: "...It will be a few weeks before it will be possible for you to read reviews in this product category."

b. Nope. They review soap, but no books. Books are for the unclean.

c. Nah. Why bother with music? No time when you're busy scrubbing with a body brush!

EPINIONS part a part b part c

a. Indeed they do have PDR-M5 reviews, although some of the links on the page go to the PDR-M1 [a much different camera].

b. No comments on the collected writings; they have only two of his published works listed, and neither has any associated opinion. A biography is rated: Einstein - Denis Brian

c. No Storm and Stress, no John Zorn, and a scant few Coltrane reviews like this rather forthright one that sticks bias up in the air and punts it.

Conclusion?

Phew. That was a bit long and tedious, but what have I learned? It is my firm belief that aggregated opinion sites that try to be all things to all people will never be detailed enough to serve folks like me. I would have much better luck finding reviews and related things on a, b, and c if I had taken my search party to sites specific to those areas. Behold:

a. Looking to the Digital Camera resource page, an in-depth review of the PDR-M5 complete with pros and cons, pictures, sample photos, and final thoughts. I'd get the same thing from most any digicam site; Steve has a review as well, with an exhaustive specs table and everything.

b. The Einstein Papers Project has a good little reference on how the volumes in the series fit together logically and chronologically. Now to agonize over which one to save me pennies for.

c. Storm and Stress review in the noisy rock zine AA.
I can decide what Zorn disc to purchase next based on all the links in Jazzweb's John Zorn section.
And I only have to look to one of the many Coltrane fan sites to decide there. Always watching my back for that beastly bias, of course.

So what's the deal? I wrote all this to hide the fact that general opinion systems are still weak? No, I firmly believe they will be improving, thanks to collaborative filtering techniques and more open opinion. Express.

That was the real conclusion, right above. From here on out, I'm tossing off an appendix of sorts. Technical articles on opinion-related subjects and methods of anonymity that wouldn't fit in the main page.

Weaving a Web of Trust
Philosophy/tech essays
Technorealism: "Information is not knowledge."

Stay anonymous on the web. It's downright scary how some of the ad agencies and other behemoths track you through your day, if you let them...

Stick it to the man! Don't stare at ad banners. I've tried the schemes of David Whitmarsh and The Great @man, but nothing beats the JunkBuster in my opinion. Be sure to also check out the modified Junkbuster that allows you to serve up blank images instead of broken ones.

I couldn't find any anonymous web gateways that were currently actually working at the time of this broadcast, but check this linklist if you absolutely must not let "them" know your IP.

Anonymous remailer linklist [thanks Nat]