In this research we introduce a new approach for Arabic word disambiguation by utilizing Wikipedia as the lexical resource for disambiguation. The nearest context for an ambiguous word is selected using Vector Space Model and cosine similarity between the word’s context and the retrieved senses from Wikipedia. Three experiments have been conducted to evaluate the proposed approach, two experiments use the first retrieved sentence for each sense from Wikipedia but they use different Vector Space Model while the third experiment use the first paragraph for the retrieved sense from Wikipedia. The experiments show that using the first retrieved paragraph is better than the first retrieved sentence and the use of Tf-Idf VSM is better than using raw frequency VSM.
展开▼
