Improved Semantic Similarity Algorithm Based on Ontology
Jun Wei, Pei Si Zhong, Chun Fen Guo · Applied Mechanics and Materials · 2011
The measuring of semantic similarity is the key technique in information retrieval. Based on the common semantic similarity algorithm between concepts, this paper makes use of semantic weighted distance and introduces such factors as node density, concept attribute and concept information content, and presents an improved semantic similarity algorithm to make the measuring semantic similarity more accurate. Through experimental calculation and analysis, the semantic similarity obtained from the improved algorithm can describe more accurately the similar relation between ontology concepts, and correspond to a higher degree with human subjective judgment.