Reputation: 4750
I have a list of words L1, and have candidate lists like B1, B2..,Bn all the lists are of the same length. But they don't have the same elements. Lists doesn't contain repeated elements. L1 is most similar to only one of B1,B2...Bn. what kind of algorithms that I can apply to find the most correlated list to L1?
Some potential algorithms are edit distances, rank correlation (which should have same elements on both lists which is not the case in mine), dice and jaccard. I'm also not sure of the word 'correlation' here is it valid in this case?
EDIT: I have a set of documents and 2 topic modelling algorithms A1 and A2. I get 10 topics by applying A1 onto the set of documents and another 10 topics by applying A2. So the above lists define the words in these topics. So now I want to find similar topics created by these 2 algorithms A1 and A2.
Upvotes: 1
Views: 950