For the last 24 hours we have been parsing Dave's new category items from his RSS feed within our k-collector w4 test server.

We have decided to split the category nesting and to match each category level to what we consider a topic. While topics should be unique and absolute within a cloud, nesting is subjective and can be defined both by the author and by the reader.

As an example, in Dave's taxonomy you can find:

- Politics
    - Presidential Election of 2004
        - Dean Campaign
        - Clark Campaign
        - ...

If Howard Dean would have his own taxonomy, it might contain something like:

- Howard
    - Presidential Election of 2004
        - Politics
        - Budget
        - Ideas
        - ...

Same topics, different nesting according to the different points of view.

After splitting Dave's categories we match them against our OPML topicRoll to see if any of our active users has already used that topic and we archive Dave's posts according to the matches we foun (guess what? a lot of categories Dave has been using already exist on our server: it's a small world ;-).

For example you can see here a post by Dave about Google which has been archived under the Search Engines topic thanks to Dave's category.

Topics on this test server require a classification (Person, Thing, Company or Who, What, Where) so we cannot automatically create new topics from a feed using categories as we currently do with ENT feeds.

But anyway it's a quite interesting first step.