|
ANNOTATION:
Second order or more tenuous associations
After running the experiment for a time to discover what are the strongest associations among words and concepts, the learning curve will taper off. More can be learned by re-running the experiment with the instruction to identify the *next* most closely associated word, or to rank the three most closely associated words in order of the strength of their relatedness. This may catch some more tenuous associations that would be altogether missed by the initial method. (Of course, a similar effect could be derived by simply varying the lists of words to choose from. This feature may have been incorporated into the experiment already.)
What name(s) would come up if we all asked ourselves, 'Who would I want to vote for for President?' Could such a learning engine help us identify those persons who embody exceptional qualities of leadership and wisdom?
Copyright© 1998 Principia Cybernetica -
Referencing this page
|
|
|
