Morphological Analysis of Tunisian Dialect

International Joint Conference on Natural Language Processing, pages 992–996, Nagoya, Japan, 14-18 October 2013. Morphological Analysis of Tunisian Dialect Inès Zribi ANLP Research group, MIRACL Lab., University of Sfax, Tunisia ineszribi@gmail.com Mariem Ellouze Khemakhem ANLP Research group, MIRACL Lab., University of Sfax, Tunisia mariem.ellouze@planet.tn Lamia Hadrich Belguith ANLP Research group, MIRACL Lab., University of Sfax, Tunisia l.belguith@fsegs.rnu.tn Abstract In this paper, we address the problem of the morphological analysis of an Arabic dialect. We propose a method to adapt an Arabic mor- phological analyzer for the Tunisian dialect (TD). In order to do that, we create a lexicon for the TD. The creation of the lexicon is done in two steps. The first step consists in adapting a Modern Standard Arabic (MSA) lexicon. We adapted a list of MSA derivation patterns to TD. The second step consists in improving the resulting lists of patterns and roots by using TD specific roots and patterns. The proposed method has been tested and has achieved an F- measure performance of 88%. 1 Introduction The Arabic Dialect (AD) is a collection of spoken varieties of Arabic. It is used in everyday communication. So, it is so important to consider it in the new technologies like dialogue systems, telephone applications, etc. (Zribi et al., 2013). The majority of these applications need a morphological analysis to segment words and to exploit their morphological features. Many important works have focused on the morphological analysis of the Arabic language, mainly on Modern Standard Arabic (MSA). AD has not received much attention due to the lack of dialectal tools and resources (Duh and Kirchhoff, 2005). However, there are differences between MSA and AD, they are considered as two related languages. Therefore, we suggest in this paper to exploit and adapt an MSA morphological analyzer (MA) to Tunisian Dialect (TD). The adaptation is done in two steps. The first step is to adapt an MSA lexicon to TD and to improve the resulting lexicon with TD specific roots and derivation patterns 1. The second step is to integrate the 1 The Arabic derivation system consists to use a set of pat- terns and roots to generate words. To generate the word “ بُ تِ كْ يَ , yaktibu, he write”, we replace the ri letters in the resulting lexicon into the MSA morphological analyzer. The paper has 5 main sections. Section 2 presents a lexical study of the TD. We present in section 3 an overview of previous works. We describe in section 4 our method for adapting MSA resources to TD. In section 5, we give the results of the system evaluation, and finally, we discuss some analysis errors. 2 TD lexical study TD is characterized by a phonology, a morphology, a syntax and a lexicon which have differences and similarities compared to MSA and even to other Arabic dialects (Zribi et al., 2013). There are many regional varieties. In this paper, we focus on the standard TD (the dialect used in the media that is the most understood …

Content truncated to 500 words