I finally got around to doing some non-work coding after going to Google Developer Day 2007 (in London). One of the big themes was ‘mashups’, which have become pretty common. Anyhow, I’ve never been much of a web developer, but I’m a huge user of Google Reader and probably add at least 1 feed a day. Anyhow, while I was there I remembered that I’ve been wanting to “classify” the data in them so I can group them together, maybe pull some trends out.
I’m not really anywhere near that yet, but I’ve got a “feed analysis” toy up on my site. When you’re there you’ll quickly see how much I’m *not* a web developer. You can just enter the URL to an RSS feed (by default it’s just the feed from my blog) and it’ll try to extract the “interesting” phrases and sentences from it. I’m not quite sure where I want to go with this, but I hope to keep tinkering with it enough that it becomes something useful for me. I believe there are already similar tools out there, but it’s still fun to play with.
Would love to hear from anybody on what types of features you’re looking for in terms of feed grouping/trending. Right now it can really only serve as a sort of ‘auto-tagging’ tool. I’m interested in document classification, but am fairly new to it, so as my naivety goes down, so should the tool.