similarity metrics based on sentence-transformers, lexical overlapping and kgtk similarity score with wikidata.