Arabic term extraction using combined approach on Islamic document
Ali Mashaan Abed, Sabrina Binti Tiun, Mohammed Albared · 2013
While a wide range of methods has been conducted to English terminology extraction, relatively few studies have been applied to Arabic terms extractio n in Islamic corpus. In this paper, we present an e fficient approach for automatic extraction of Arabic Termino logy (SWTs, MWTs). The approach relies on two main filtering steps: the linguistic filter, where simple part of speech (POS) tagger is used to extra ct candidate MWTs matching given syntactic patterns, and the statistical filter where several statistical methods (PMI, Kappa, CHI-squire, T-test, Piatersky- Shapiro and Rank Aggregation) are used to rank candidate MWTs and we applied IF.IDF to rank the SWTs candidate. Our approach extracted the bi-gram candidates of MWTs Islamic term from corpus and evaluated the association measures (STWs and MWTs) by using the n-best evaluation method .