Wednesday, March 21, 2012

Artificial Intelligence, Wikipedia stub pages, Cutting Edge of Information, paraphrasing, machine translation

 
When I enter a query into a search engine, I envision  the closest intersection of the search terms to be the domain of the answer. This does not require the search terms to be present on result answer page. This is not a natural language query but a more specific implementation of pigeon speak. Most of my blog posts including this one fits that model.

I imagine a reference similar to Wikipedia. call it  "wishipedia" where the result answer would be auto generated at first arrival if the page does not exist, but allow algorithmic and user edits. In a single keyword scenario then there would be a 1 to 1 relationship with that Wikipedia wiki page. 

This would be a valuable resource at the "cutting edge of information". Think of immature Wikipedia stub pages as seeds. When a user arrives to a stub page, little or no real information is present at Wikipedia. Since the page is mostly blank, this discourages user participation. Wishipedia  would provide a little more structure with a template similar to a "fill in the blank question." This should lower the barrier to enter information.  This would also invoke more game dynamics.

The key here is not to replace Wikipedia but to give more traction to information sitting on the cutting edge. This Wishipedia page can then be fed back into Wikipedia and follow the normal review process.

Similar to the book a new kind of science this would be "a new kind of knowledge" More fuzzy, less clear, a little erroneous, but another small step yielding complex patterns toward the future of Artificial Intelligence.  

No comments:

Post a Comment