A Comparison of Three Machine Learning Methods for Amazigh POS Tagging

Part of speech tagging (POS tagging) has a crucial role in different fields of natural language processing (NLP) including Speech Recognition, Natural Language Parsing, Information Retrieval and Multi Words Term Extraction. This paper describes a set of experiments involving the application of three state-of the-art part-of-speech taggers to Amazigh texts, using a tagset of 28 tags. The taggers showed encourageous performance, in particular having problems with unknown words. The best results were obtained using a decision tree approach, whille CRF and SVM based taggers got comparable results.