Pittsburgh's online weekly Pop City has an article today about interesting Pittsburgh startup mSpoke, working "at the hot intersection of the semantic and implicit web, an area that involves the synthesis of personal web data." Try mSpoke’s FeedHub personalization technology for RSS feeds. It's very easy to use. Just upload the OPML from you feedreader (if you use Google Reader you can easily export the OPML for all your feeds). Then add the FeedHub feedreader to your homepage on Google or Google Reader. The cool thing is that it looks at all the feeds you subscribe to (for example, mine include Jurist, LawLibTech, Legal Scholarship Blog, Madisonian, Wired, CNet, the WSJ law blog, and other law/technology types). FeedHub will then "feed" you information by choosing great posts based on what it has learned from your OPML.
FeedHub chooses personalized content for you based on both your reading behavior and explicit feedback. The more posts you read, and the more you tell FeedHub how you feel about your content, the better we'll get at choosing content to feed you.