Hi All,
I‘m looking for an approach for evaluating results of topic identification. I applied two different methods on text dataset to extract topics. The results are list of topics, each topic have list of terms. Actually, the standard evaluation is performed by assigning each document into related topic. Here I’m not interesting on documents, I’m only interesting on the terms are occurring together in each topic.
Result 1
Topic 1 : term1, term1 ,,….. termN
..
Topic N : term1, term1 ,,….. termN
Result 2
Topic 1 : term1, term1 ,,….. termN
..
Topic N : term1, term1 ,,….. termN
Does anyone knows an approch to comapre the Result 1 and Result 2?
Thanks
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts
More Donni Darsh's questions See All
Similar questions and discussions