Memory-based morphological analysis and part-of-speech tagging of Arabic

Memory-based learning has been successfully applied to morphological analysis and part-ofspeech tagging in Western and Eastern-European languages (Daelemans et al., 1996; Van den Bosch and Daelemans, 1999; Zavrel and Daelemans, 1999). With the release of the Arabic Treebank by the Linguistic Data Consortium, a large corpus has become available for Arabic that can act as training material for machine-learning algorithms. The data facilitates machine-learned part-of-speech taggers, tokenizers, and shallow parsing units such as chunkers (Diab, Hacioglu, and Jurafsky, 2004); cf. Chapter 9. However, as argued and illustrated throughout this book, Arabic offers special challenges for data-driven and knowledge-based approaches alike. An Arabic word may be composed of a stem consisting of a consonantal root and a pattern, and may furthermore contain affixes and clitics. Arabic verbs, for instance, can be conjugated according to one of the traditionally recognized patterns. There are 15 triliteral forms, of which at least 9 are common. They represent very subtle differences. Within each conjugation pattern, an entire paradigm is found: two tenses (perfect and imperfect), two voices (active and passive) and five moods (indicative, subjunctive, jussive, imperative, and energetic). Arabic nouns show a comparably rich and complex morphological structure. In this chapter we explore the use of memory-based learning for morphological analysis and part-of-speech (POS) tagging of written Arabic. The next section summarizes the principles of memory-based learning. Section 3 describes the data used throughout the study for both tasks. The subsequent three sections describe our work on memory-based morphological analysis (Section 4) and its integration with part-of-speech tagging (Section 5). The final Section 6 contains a short discussion of related work and offers an overall conclusion.