Highlighting relevant concepts from Topic Signatures

摘要: This paper presents deepKnowNet, a new fully automatic method for building highly dense and accurate knowledge bases from existing semantic resources. Basically, the applies knowledge-based Word Sense Disambiguation algorithm to assign most appropriate WordNet sense large sets of topically related words acquired web, named TSWEB. is personalized PageRank implemented in UKB. improves by means current content creating volumes semantic relations between synsets. KnowNet was our first attempt towards acquisition relations. However, had some limitations that have been overcomed with deepKnowNet. deepKnowNet disambiguates hundred all Topic Signatures web (TSWEB). In this case, highlights relevant word senses each Signature filter out ones are not so topic. fact, the it contains outperforms any other resource when empirically evaluated common framework based on similarity task annotated human judgements