I do research on automatic opinion identification, and one thing that is really interesting to the community right now is analyzing blog data. Most of the available tagged resources are over newspaper data or movie reviews or other kinds of collectable TEXT – user star rating type things (restaurant reviews, product reviews, etc.)

The community is very interested in moving to blog data, where ostensibly there would be more and varied opinions available to analyze, but there isn’t too much data available for that yet. (But see the TREC Blog track work where there is an opinionated blog search task.)

All this interest in blogs and user generated media seems to have had an impact on “traditional” print media. Recently, there was a workshop on the Future of News. It was held at Princeton University, near my old stomping grounds. It would have been nice to go, but thanks to the (news-media destroying!) blogs, I’ve been able to at least get a brief impression of what was discussed.

Matthew Hurst’s great Data Mining blog has two posts with pointers to some summaries from the workshop. Fun stuff.

I’m firmly of the opinion that traditional news media will be around for a long time. Blogs do have some role to play in modern news dissemination, but not a large enough role to displace focused organizations that can fund people to do research and have a vested interest in vetting information. It isn’t clear to me that the newswires do as much of this as they should, but the traditional media certainly will play a role in choosing what news to elevate to the national level.

At some point, most blogs are really locally focused, and I don’t see how any of the personally-run small sites (like, say, my blog) could ever hope to break interesting news more than once in a lifetime. Also, I like writing about what I ate for dinner. That isn’t news. 🙂