I built something like this in 2005[1], got distracted[2] and still wish I had the tool today. My database was organized around links rather than feeds/items. T

I built something like this in 2005[1], got distracted[2] and still wish I had the tool today. My database was organized around links rather than feeds/items. Two stories that included links to the same thing were considered to be talking about the same thing and would be presented together as one part of the conversation. This was a low rent (but more accurate) version of what Google News does where they assume two articles are about the same subject by comparing the terms used in the articles. I also used Bayesian classification to determine which articles were ones you were more likely to be interested in reading. [1]: the marketing pages are still online for this ancient piece of code: http://blazingthings.com/zestynews/tour1.html http://blazingthings.com/zestynews/tour1.html [2]: I created the TurboGears web framework while creating Zesty News, and ultimately decided to pursue the framework rather than Zesty News.