Resolving Malay Word Sense Disambiguation Utilizing Cross-Language Learning Sources Approach

摘要

Ambiguity in Malay words is a huge issue in interpreting Malay language to different languages. Authors propose a Malay Word Sense Disambiguation (MalayWSD) to resolve ambiguous Malay words using cross-language learning sources such as AsianWordNet (AWN) and PrincetonWordNet (PWN).The MalayWSD is a semi-automated on non-particular area utilizing disambiguation systems, word structures and window sizes. The disambiguation methods incorporate vector, vector-pair, path and lesk. In this paper Malay words that have different meanings and part-of-speech are experimented.The results reveal that utilizing AWN with vector method and PWN with path procedure gives best exactness. In any case, for MalayWSD the time execution, the vector procedure with AWN at five window dimensions are acceptable.