A Hidden Markov Model -Based POS Tagger for Arabic
Fatma Al Shamsi, Ahmed Guessoum · 2006
This paper presents a Part-of-Speech (POS) Tagger for Arabic. The POS tagger resolves Arabic text POS tagging ambiguity through the use of a statistical language model developed from Arabic corpus as a Hidden Markov Model (HMM). The paper presents the characteristics of the Arabic language and the POS tag set that has been selected. It then introduces the methodology followed to develop the HMM for Arabic. The proposed HMM POS tagger has been tested and has achieved a state-of-the-art performance of 97%.