Collaborative Filtering dot ORG

co +  labor + felt = working together to filter [Reality?]


Mission Draft: to increase the quality of the knowledge and information with which we make decisions.

What could this possible mean?

Correlation - look for the same pattern in different places. Big, common
themes may reinforce and stand out while others cancel each other out.
Could be over time over space (people). A very powerful variation of
this is across domains. Seeing the correlation between the work in
thermodynamics and in graphics could lead you to radiosity.

Distance - looking at something closer or farther away can allow us to
see patterns that we couldn't see before. The classic "forest for the
trees" problem.

Filter noise out - removing data out of scale can make the valuable data
pop out. For instance, dropping the high frequency, periodic noise from
an image can clear it up.

Perspective - similar to distance in some ways, the ability to change
your point of view can have a big impact on what you see.

Other Links:

Berkeley's School of Info Management's Collaborative Filtering page.

A Joint Effort of John Sosoka & Jan Hardenbergh Have ideas? Send eMail to YON

created 26-AUG-2000