Slashdot Mirror
Microsoft Tracking Behavior of Newsgroup Posters
theodp writes "Ever get the feeling your Usenet newsgroup list is being watched? By Microsoft? If so, consider yourself right. An interesting but troubling CNET interview with Microsoft's in-house sociologist goes into how the software giant is keeping a close eye on newsgroups and other public e-mail lists, tracking and rating contributors' social habits and determining "people who the system has shown to have value." Those concerned that it's not a good idea for computers to track their belongings and whereabouts are advised that they may ultimately have to fragment their identities, keeping multiple IDs and e-mail addresses."
" Those concerned that it's not a good idea for computers to track their belongings and whereabouts are advised that they may ultimately have to fragment their identities, keeping multiple IDs and e-mail addresses."
Who isn't already doing this?
With the advent of spam most people I know abandonned their first email address years ago. I have one for each service I use (including slashdot).
The AURA just sounds like the CueCat Digital Convergence people who wanted to put a bar code on everything. Again, MS is not the company I'd like to see doing this.
*Rather Offtopic - but Digital Convergence used to advertise the CueCat with an 'Angel coming down to earth from heaven to barcode everything' and the well-known Digital Angel RFID people have also made a religious reference in the company's name. The hue and cry of Christian's 'the number of the beast' references beg the question:
Who the hell is doing marketing for these people? I remember getting an icky feeling when I saw the 'infomercial' for the CueCat, and similarly the Digital Angel website. And I'm not the 'churchy' type. I can only imagine what the fundies think...
* This idea is copyrighted. Use of this idea may not be used to more attractively market 'evil' technology, or put a chip in my head. Thanks.
How is this "troubling"? They are researching a way to make USENET and such more effective. They aren't interested in the fact that cmdrtaco@slashdot.org posted to alt.sex.unicorns 10 times last month.
This is good valid research, the type that applied research CS programs should be doing. Thismay actually make a difference in a deployed product.
I think we should tone done the M$ and SCO crap for a while.
The next day he was showing Ben Schneiderman some of this stuff at the open house. A bunch of us looked on as they chatted, planned visits, golf outings and talked about how it all worked.
Depending on the queries he gave it, this one program would chew through data from usenet. and give back all kinds of stats and then draw relationships It even did graphical representaitons of users' actvity. Density of posts in a single thread versus starting new threads, frequency of posts, replies vs. new messages etc would be denoted by distance from the main timeline, darkness and width of the circel and so forth. You would look at a wide but faint circle and say (and I may be off in how the key worked, but ...) "This guy sticks to the topic over a long period of time" or you could denote the flame warrior or the vagrant by their graphical representation and so forth. The way the data was processed was really cool and how quickly you could start to decipher the keys was really interesting.
The Big brother implications ... well that's a whole 'nother thing there too isn't it?
Que Deus te de em dobro o que me desejas
[May God give you double that which you wish for me]
One of the first tasks of any individual joining a group is to determine the pecking order within which authority is distributed. This is a critical task that humans have been doing since before they were human.
What's being talked about here is reverse engineering trust heirarchies, algorithmically, simply from a discussion corpus extracted from Usenet.
This is very, very cool stuff. It is a hard application of a soft science, and if its results match empirical data, it represents a greater level of understanding about the human mind.
This is something to celebrate and take interest in, not malign simply because it's Microsoft that's behind it.
I do remind the security paranoid that reputation management remains one of the few characteristics obsessively protected in otherwise anonymous systems.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
While it's true that such simple messages and "Fr1st p0st" equivalents can't be mapped, they can easily be filtered out as noise, since there's no content anyway.
However, if you're posting reviews to Amazon or ePinions your text is likely to have analyzable content.
I know someone who has done this type of analysis and discovered people who reply to their own posts in dicussion boards under different IDs to make it look like they had some kind of consensus. When confronted with the analysis, they admitted the ruse.
At least on their own newsgroups (the microsoft.* hierarchy) they've been doing this for years. Back over 6 years ago I was a Windows programmer (don't worry, a long time ago I saw the light and now am a linux programer!) Anyway... Because of the work I was doing at the time with Windows and was also answering a lot of questions in the microsoft.* newsgroups I attracted MS's attention. They made me an MVP (Most Valuable Professional) back around '95.
One of the things MVP's were told was that MS tracked our posting habits in their newsgroups. They used our e-mail addresses for this. The tracking was purportedly to help determine if our MVP status would be retained from year to year. (it's an annual award) Since they acknowledged way back when that they were tracking users on their own newsgroups it really doesn't surprise me all that much that they'd expand it to cover more groups.
Actually, given that Google has an archive of many of the newsgroups it really wouldn't be all that difficult for pretty much anybody do track individual posting habits, etc. Just run some searches for the e-mail address of the user in question.
For a few months I did this for MS Games. I searched newsgroups and fan message boards to see what players were talking about. If everyone was pissed off and wanted a new patch, that's what I reported. If people were excited about a certain feature, I reported that, too. If you ever flamed MS for something you didn't like, I might have sent it to them.
One choice quote from memory... "WE NEED A PATCH. GOD IF YOU SHOVED SOME COAL UP THERE ASSES YOUD GET A DIAMOND!!!LOL"
It paid $10/hr, and I needed the money.
I haven't used newsgroups much, and therefore my opinion may be inaccurate, but it seems like anyone looking an groups using software with theses new search features is going to approach things very differently than people using tradition methods. Essentially, if there if a group can be called a community, it's probably that way because everyone who spends much time there knows each other (to some extent), follows whatever have developed, and so on. Someone who comes in because their news reader told them the group was popular is not going to see any of that, and if it happened too often then it could be rather disruptive. If it happened a lot, then it could change the way people handle themselves in the group and the methods used to rate threads and authors might become useless.
It also seems like the ratings could concentrate posts too much. If people use the system to search old threads then it wouldn't be an issue, but if it gets used to find places to ask things then it could increase the number of questions that could have been answered by searching, RTFMing, etc.. If only the best resources get used, then they could grow to the point of becoming impossible to search while everything else is ignored.
Finally, I wonder about the good posters as a support resource attitude. Obviously plenty of people are willing to help others online, but that doesn't necessarily mean they want everyone coming to them for assistance. Again, it wouldn't be a problem if the system encouraged searching only resources, but if it ended up encouraging un-researched posts then it could flood good groups and authors with unnecessary questions. (Obviously some answerers are going to be fine with those questions, but in bulk they tend to get annoying.)
None of these things are necessarily an issue at all, of course -- they would only be a problem in the context of Microsoft releasing a news reader with their search features (as was implied by the last article on the topic) and getting a lot of people to using. If it remained a search tool that wasn't used all the time then it could be very useful.
I was wondering why reading the aritcle made me feel squeamish. It's the thumb on the scale. No matter what assurances anyone makes, that thumb will be on the scales. In contrast, Slashdot comes off as professional, essentially regardless of what they do or how well they do it ( or how badly they mangle it ;-)