Unified Guidelines and Resources for Arabic Dialect Orthography

Unified Guidelines and Resources for Arabic Dialect Orthography Nizar Habash, Fadhl Eryani, Salam Khalifa, Owen Rambow,† Dana Abdulrahim,‡ Alexander Erdmann, Reem Faraj,† Wajdi Zaghouani,∗ Houda Bouamor,♣ Nasser Zalmout, Sara Hassan, Faisal Al-Shargi,⋆ Sakhar Alkhereyf,† Basma Abdulkareem,† Ramy Eskander,† Mohammad Salameh,♣ Hind Saddiki New York University Abu Dhabi, UAE, † Columbia University, USA, ‡University of Bahrain, Bahrain, ∗Hamad Bin Khalifa University, Qatar, ♣Carnegie Mellon University in Qatar, Qatar, ⋆Univisität Leipzig, Germany nizar.habash@nyu.edu, rambow@cs.columbia.edu, darahim@uob.edu.bh, wzaghouani@hbku.edu.qa, hbouamor@cmu.edu, alshargi@informatik.uni-leipzig.de Abstract We present a unified set of guidelines and resources for conventional orthography of dialectal Arabic. While Standard Arabic has well defined orthographic standards, none of the Arabic dialects do today. Previous efforts on conventionalizing the dialectal orthography have focused on specific dialects and made often ad hoc decisions. In this work, we present a common set of guidelines and meta-guidelines and apply them to 28 Arab city dialects from Rabat to Muscat. These guidelines and their connected resources are being used by three large Arabic dialect processing projects in three universities. Keywords: Dialectal Arabic, Orthography, Phonology, Morphology, Conventions 1. Introduction Arabic dialects are linguistic varieties that are historically related to classical Arabic and co-exist with Modern Stan- dard Arabic (MSA) in a diglossic relationship. While MSA, the official language of all Arab countries, has well-defined orthographic standards, Arabic dialects have no official or- thographies. As such, besides unintentional typographic er- rors, no spelling of a dialectal word can be considered “in- correct.” And since Arabic dialects vary from MSA and from each other in terms of phonology, morphology, lex- icon and syntax (Watson, 2007), using MSA orthographic standards cannot fully address the needs of the dialects. As an example of the degree of variety in dialectal spelling, Figure 1. presents the 27 actually attested spellings of one Egyptian Arabic word online. The large number of possi- bilities results from independent decisions such as whether the proclitic /ma/ should be written attached or separated (+� m+1 or �� mA), or whether to write the stem in a way that reflects its phonology (��� ˆw), or etymology ( �� q). Habash et al. (2012) introduced the concept of Con- ventional Orthography for Dialectal Arabic (CODA); and they proposed a set of guidelines and exception lists for Egyptian Arabic. Their conventions were used in the Lin- guistic Data Consortium for annotating Egyptian Arabic (Maamouri et al., 2014). Since then, a number of additional efforts followed suit for other dialects (Zribi et al., 2014; Saadane and Habash, 2015; Jarrar et al., 2016; Khalifa et al., 2016). While the original CODA guidelines aimed at being easy to adjust to new dialects and contained some 1Arabic script transliteration is presented in the Habash-Soudi- Buckwalter transliteration scheme (Habash et al., 2007): �� �� �� �� �� � � � ��� �� � �� � �� � �� � �� �� �� � � � �� � � ��  b t θ j H x d ð r z s š S D T ˇD ς γ f q k …

Content truncated to 500 words