Two grad students at U of Toronto have been working this term on a tool to
analyze chat conversations (IRC, IM, etc.) and cluster messages into
threads. They presented their work this afternoon, and will be doing
another presentation on April 9; I think it has promise, especially since
they're using incremental algorithms that would be well-suited to
real-time execution. They're not going to carry on with it (they have
thesis topics picked out, and this ain't it :-), but the long-term goal
would be to build something that could plug into various clients, or be
used to process logs.
Would this make a good project for XMPP SoC? It would require a student
who already has some understanding of IR, clustering algorithms, and NLP,
but it could help make large chat spaces more usable. I'd be willing to
mentor...
Thanks,
Greg Wilson
http://www.cs.toronto.edu/~gvwilson
