A Hybrid Approach to Discover Entity Synonyms

Mamta Kathuria, Chander Nagpal, Neelam Duhan

OAI: oai:igi-global.com:300296 • DOI: 10.4018/IJIRR.300296

A major part of the current web information retrieval involves the fulfilment of people's daily needs based upon finding entities involving movies, events, persons, places, concepts, etc. The trend has been the outcome of the tremendous amount of precise and detailed information available on the web making it possible to retrieve the precisely targeted information involving specific instance of a particular domain called entity. The major problem in the entity-based search is the diversity of the name references made to the same instance of the same domain by the content creators and the searchers. These diversified references are known as entity synonyms. The problem is that they cannot be handled through lexical resources. This paper takes into account the previous works carried out in this direction and proposes an advanced mechanism that creates a richer set of entity synonyms than the previous approaches. The work also proposes an index to assess the quality of entity synonyms so generated which is further normalized and fuzzified for implementing automated search.