The documentation for cts:similar-query() states that it "uses an algorithm which finds the 16 most "relevant" terms in the model nodes". Solr's MoreLikeThis handler is similar, but there the mlt.interestingTerms parameter allows you to retrieve the list of "interesting/relevant terms" (those with a high TF-IDF score) used to build the 'similar documents' query - is this possible in MarkLogic?
alf _______________________________________________ General mailing list [email protected] http://xqzone.com/mailman/listinfo/general
