Cluster list of comments

Question

I am a beginner who just has discovered the great Carrot2 framework .

I try to use the Java API to cluster a list of Facebook comments (~100 comments with about 10-200 characters) with the LingoClusteringAlgorithm. Can I use the comment text as "title" field and leave the fields "snippet", "url" and "query" empty? Or is there a better way (f.ex. using the comment text twice for "title" and "snippet"?)

Answer 1

At least one of "title" or "snippet" must be non empty, so you can put the post text in the "snippet" and leave "title" blank. It shouldn't matter whether you leave "title" or "snippet" blank, the results should be the same in both cases.

The "url" field is used only for display purposes, it does not affect the results of clustering. You can leave it empty or put the direct link to the post if you plan to use it in the UI.