0

I write a LPA algorithm to detect community, but when I try to construct the graph, I find that the edges is very huge (suppose 100K pepole see 1 movie, then I got 10^6 * 10^6 edges more than 100 billion), what should I do to deal with this, or JUST filtering the big collections?

gavinzjchao
  • 1
  • 1
  • i would not use cfml to solve this problem. instead i would build a service in C to do the heavy lifting then call this service with cfml. That way you can make much better use of memory and cpu resources. we often use this approach. with cfml just being the glue that ties a bunch of business logic together as a REST backend or directly to render a HTML UI. – M.Scherzer Mar 15 '16 at 11:33
  • A service for computing the pair distance? I don't understand your solution , Could you give me some examples for your ideas? thanks! – gavinzjchao Mar 17 '16 at 11:53
  • See this question for references, including a Java library that may solve your problem. http://stackoverflow.com/questions/5822265/are-there-implementations-of-algorithms-for-community-detection-in-graphs – Robert Munn Mar 25 '16 at 07:19

0 Answers0