Semantic Similarity Measures: an Overview and Comparison

Dr.B. Poorna, A. Sudha Ramkumar,


Similarity measure is calculated based on the syntactical representation of terms. Similarity measure used in data mining task likes clustering, and classification returns irrelevant information. The semantic similarity calculated based on the relatedness between wordpairs of terms returns better result. Many researchers proposed approaches for getting word similarity by using different sources like ontologies, thesauri etc. This Paper provides an overview of six existing semantic similarity measures and compared those semantic similarity measures using the wordpairs of sports domain ontology. This paper describes about how WordNet is used to retrieve the synonyms and using synsets how semantic similarity measures are calculated. Finally, comparison of selected semantic similarity measures for the given wordpairs with respect to the considered knowledge base domain ontology and WordNet is presented.


Similarity measure; WordNet; Synonyms;

Full Text:




  • There are currently no refbacks.

Copyright (c) 2018 International Journal of Advanced Research in Computer Science