***** To join INSNA, visit http://www.sfu.ca/~insna/ *****
You have a situation where many papers share one keyword and few people
share another keyword. You seem to want to treat the rarer ties so created
as "stronger". It is not apparent why you would want that. If these are
ties created by working in the same area, then there are more ties among
the more people in the popular areas than among the few people in the
unpopular areas. I don't think that the former ties need necessarily to be
weaker than the latter ties. If you think that more ties necessarily means
weaker, then you should theoretically specify how strength declines with
number of ties and why. Your theory should then dictate how you weight
accordingly. You might instead treat this as an empirical issue, IF you
have some other indicator of tie strength.
At 03:06 PM 4/8/2004 -0400, Jim Moody wrote:
>***** To join INSNA, visit http://www.sfu.ca/~insna/ *****
>Hi Folk -
>Hope you are well.
>I'm working with a student who is creating a network based on overlapping
>terms (call them keywords for simplicity), and we're looking for a way to
>scale the overlap by the relative frequency of the words.
>So, for example, you might have two papers that share 3 keywords, but those
>keywords are very common, compared to a paper where they share 3 very rare
>keywords. We would want to weight the link between the two papers in the
>second case higher than the first.
>Any suggestions would be appreciated.
>Department of Sociology
>Ohio State University
>372 Bricker Hall
>190 N. Oval Mall
>Columbus, OH 43210
>[log in to unmask]
>SOCNET is a service of INSNA, the professional association for social
>network researchers (http://www.sfu.ca/~insna/). To unsubscribe, send
>an email message to [log in to unmask] containing the line
>UNSUBSCRIBE SOCNET in the body of the message.
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.sfu.ca/~insna/). To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.