Talk of attention data has been a hot topic among web mucky mucks for some years now, though as of yet, not much has been done other than talk. Attention data for those that are forgivably behind the curve here, is simply information about what you're doing, how you're doing it, and how long you've been doing it. For example, you might record that you spent an hour reading feeds in Google Reader this morning, which of them you opened, how long you spent reading each one and then go on to record that you spent 10mins talking in IM with a friend. You might go as far as to record what you spoke about, and if you were busy doing other things at the same time as that conversation. Attention, with me? Ok good. So, IAN, a German advertising company operating on over 50 islands within Second Life and already working on behavioral targeting projects, are developing a set of tools to record attention in the Metaverse, with the purpose of anonymizing that data, and buying it back off of residents to better tune advertisements inworld.
Sound a bit scary? It is. The thing about attention, and behavior is that in order to be effective, it has to be very personal. Not everyone is comfortable having their behavior recorded, and that's probably the biggest stumbling block for what could be a killer personalization structure that could theoretically span the Metaverse, the Web, and everything you do on your computer, or even cell phone. The benefits of such a system could be massive, in fact they could change the way we live, work and play forever, but as of yet, it's still all a lot of hot air and breathless chatter from the muckies.
Why would it benefit you? Imagine being able to export your attention data from your browser, and build an "attention profile" that could in turn be imported into Second Life for example. That data could be used to help you find information relevant to your interests, meet others with similar profiles, tune advertisements so they actually gave you information you wanted to receive, anything.. The possibilities are endless. And this is what IAN are working toward. Sebastian Kupers emailed me to say that his firm had joined the APML working group (Attention Profiling Markup Language) and were actively trying to bring attention data tools to the Metaverse, as well as add a voice from the 3D web to the group. Though the plan is ambitious, it's not the technical implementation that's the issue, most of that is fairly simple, it's trust that poses the biggest obstacle. Second Life does not support cookies, meaning that any data, or attention profiles built for individuals would have to be stored on an IAN database, which is far from ideal. Trustworthy, nice folks though they may be, many people would be a little spooked by the idea I thnk, and certainly Sebastian seems to be well aware of this.
Despite trust issues, technological hurdles and implementation challenges though, the project is a worthy one I think. Attention profiling will eventually make all of our online lives much more productive and pleasurable, and in the Metaverse, the project has to start somewhere. I have to yet again wonder though, that would a wearable HUD, (as would almost certaily be the way to deploy such tools) be effective? Integration into the viewer with an opt-in system for recording and publishing would seem a more effective way to record attention data, and a better way to aggregate it.
Hm. As somebody who works every day in the field of data mining and processing, I have to say that the longer I work, the more I fear. I know what _I_ could do with information gathered, and I'm sure that there are a million people cleverer than I working on the matter.
At the moment I am in the position of not being willing to accept any data gathering at all without a clear policy of how those data are going to be used and distributed in the future. (I'm careful enough that I have my product update servers being explicitly opt-in.) Such things are required under various bits of UK (and EC, I think) law, bt I don't see them turning up very often in practice - I've yet to read a coherent data retention statement in SL.
Yeh, it's not an easy subject. There's so much potential for good here, but it comes with an equal or greater potential for abuse..
I think potential is really huge, but the dangers are as well... but if we as a society are going to evolve into some more advanced form then we almost certainly don't have any choice - we will almost be forced to use more and more of these services.
On the other hand, it should be pretty easy to find out : which service is malicious and which is meaning honest business. The news with new social media ( for example blogs) will spread at a speed of light about any malicious attempts and for a service with that reputation it will be almost impossible to regain trust...
No trust -> no users -> no business
So again in my opinion - it's about honesty and clear set of rules, on what this gathered info will be used...and about never bending those rules.
That's exactly how Meta Mart plans to be - never bending it's rules, and always serving it's users for their benefit.