Adventures in lifestreaming and attention - part one

In recent months I have been getting more interested in the attention economy and APML. As I understand it this a way of an individual owning the attention that you give to content on the web. For a  summary of the basics of attention profiling try this excellent article by Marjolein Hoekstra.  Everything that you read, share and comment on is an investment of time and you should be able to own your data and have control over it as you use various sites and services. At present this information sits in many 'silos' for example within google and facebook. Wouldn't it be great if you could store it all somewhere centrally and make use of it?

In addition to personal interest in your own attention profile, the sites and services that you use should be able to improve what they serve up to you if they can better understand what type of content is likely to get your attention. You might even be able to sell your attention data to advertisers, or so the theory goes.

The practical benefits of storing your attention data are as yet unproven but I thought it would be an interesting experiment to see what data it would produce.

The first step was to create a lifestream, consisting of all the feeds that are about 'me' in one place, so the feed from this blog, from my RSS reader shared items, Twitter, my vox blog etc. I tried out 3 services for this;

Tumblr - which has now created a lifestream blog for me there are plenty of features on top of producing an aggregated feed and this service is well worth a look.

Meinone - which in a way is a similar service and very easy to use to create an aggregated page and Feedbite which is an easy way of creating an OPML file of your lifestream.

The second step was to feed this data (using the feedbite OPML) to a service called Engagd using their attention profiler. This service creates an APML file of your attention data, for example my APML is here.

So now I have some good sources of aggregated feeds of all my web attention and an APML file. I'm going to track it for a few weeks and see what the benefits are to me as a user and whether I can feed my attention data to various services to improve their targeting of content for me. I haven't gone as tracking my clickstream yet with a service like Cluztr but that might be a next step. More in a few weeks.....

Comments

[Chris Saad]

Look forward to hearing the results Piers

[Piers Jones]

Thanks Chris I will :) any tips on tuning your attention profile? Is it a matter of the more feeds the better it gets? Or the longer it runs the better it will be? So far Engagd has recorded 16 concepts in my APML, I'm guessing that this builds up over time?

[Chris Saad]

At the moment Engagd will continue to keep about 15-20 concepts in your implicit concepts - we will be soon expanding this to a more relative number.

The concepts get better over time, but are also improved if you use direct feeds rather than a spliced feed typically.

How are the concepts in there so far?

Also if your APML is created from public info like your lifestream - you might like to share it with others on your blog sidebar using a badge or the cluztr APML tagcloud widget.

Be sure to set it to public in Engagd so others can access it though.

[Piers Jones]

thanks Chris that helps, I have created a tag cloud for the widget. I'm still waiting on a beta invite for the full Cluztr service so if I get an account sometime soon I'll try adding in my clickstream.