IdeaCenter / IdeaFisher / IdeaBank


Here's a software package that sounds very fascinating: IdeaFisher (or is it IdeaBank?). I find this very fascinating, as my own idea of "associative text" follows very similar basic assumptions. But IdeaFisher uses a very pre-structured approach to associations:

"[T]he IdeaBank is an organized storehouse of more than 65,000 idea words, phrases, and concepts familiar to modern society. The software's cross-referencing power creates more than 775,000 direct associations."

So this database is a completly human-built cross reference of concepts. And, as is natural for a closed source commercial database, it is static.

Which got me thinking: there is a ever growing, very well cross referenced database of knowlege and concepts freely availeable: Wikipedia, you guessed it. Now how could one build a similar tool making (realtime) use of the evolving Wikipedia cross reference?

The approach of using text-similarity (statistical) would only well work for controlled, small sets of texts. DEVONthink is a very powerfull example of this school of thought, and it appears to work very nicely. But due to the indexing cost, none of these tools would work in real-time of a set as large and constantly moving as Wikipedia (or the internet at large). You could try and import Wikipedia into a tool like that... but you'd loose freshness if you didn't repeat every day or so... Only if the Wikipedia servers did that kind of indexing themselves would this be a viable approach.

But maybe using statistical text-comparison is overkill in the first place? Take a look at IdeaFisher's example session. I have a feeling that much of that might be done with Wikipedia's basic keyword search and some term extraction from the resulting pages...


<< Incremental Reading  |  Vlad, a request >>

alles Bild, Text und Tonmaterial ist © Martin Spernau, Verwendung und Reproduktion erfordert die Zustimmung des Authors

Martin Spernau
© 1994-2024 Wunschliste

Facebook me!


powered by Traumtank