JIT-IR and Offshoots

["Srikant Jakilinki" <sriks dcs gla ac uk>]

Hi

1) This is my first post and being a PhD student, I would like to  
contribute towards the theoretical side of things. When I first read about  
Dashboard, it reminded me of something called "Just in Time Information  
Retrieval" work done by Bradley James Rhodes. There are other stuff out  
there that I had collected a few weeks back when I saved the search ["just  
in time" information retrieval] on Google. Anyway, here are the main links  
-
IBM Article - http://www.research.ibm.com/journal/sj/393/part2/rhodes.html  
(quick primer)
Rhodes' Thesis - www.bradleyrhodes.com/Papers/rhodes-phd-JITIR.pdf (has  
some nice theory and background)
Emacs Plugin created from above - http://remem.org/
Autonomy's Active-Knowledge product -  
http://www.autonomy.com/Content/Products/Interfaces/Active_Knowledge (done  
about 3 years ago)
Blinkx - http://www.blinkx.com/overview.php (FYI, the founders of Blinkx  
were former employees of Autonomy)
I hope they are useful. Most importantly, I hope we are not reinventing  
some of the things and implementing some of the coolest stuff  
(functional/conceptual mimicry). On this note, it would be nice if any of  
you would give their take on how Dashboard would be different from the  
above to clear up any confusion that may result

2) I intend to do two things for the Dashboard project. Do some usability  
testing and add an information-extraction (extract persons, locations,  
dates) from the filtered text which could then be used by GeoURL and FOAF  
kind of services. If anyone of you guys have any pointers or suggestions,  
please let me know.

Cheers,
Srikant