Computing semantic relatedness using Wikipedia-based explicit semantic analysis-英文文献.pdf

Computing Semantic Relatedness using Wikipedia-based Explicit Semantic Analysis Evgeniy Gabrilovich and Shaul Markovitch Department of Computer Science Technion—Israel Institute of Technology, 32000 Haifa, Israel {gabr,shaulm}@cs.technion.ac.il Abstract We propose a novel method, called Explicit Semantic Analysis (ESA), for fine-grained semantic representation of Computing semantic relatedness of natural lan- unrestricted natural language texts. Our method represents guage texts requires access to vast amounts of meaning in a high-dimensional space of natural concepts de- common-sense and domain-specific world knowl- rived from Wikipedia (), the edge. We propose Explicit Semantic Analysis largest encyclopedia in existence. We employ text classi- (ESA), a novel method that represents the mean- fication techniques that allow us to explicitly represent the ing of texts in a high-dimensional space of concepts meaning of any text in terms of Wikipedia-based concepts. derived from Wikipedia. We use machine learning We evaluate the effectiveness of our method on automatically techniques to explicitly represent the meaning of computing the degree of semantic relatedness be