Building a Morphological Analyser for Laz

Proceedings of Recent Advances in Natural Language Processing, pages 869–877, Varna, Bulgaria, Sep 2–4, 2019. https://doi.org/10.26615/978-954-452-056-4_101 869 Building a Morphological Analyser for Laz Esra Önal Cognitive Science, Boğaziçi University Istanbul, Turkey esra.onal@boun.edu.tr Francis M. Tyers Department of Linguistics Indiana University Bloomington, United States ftyers@iu.edu Abstract This study is an attempt to contribute to documentation and revitalization efforts of endangered Laz language, a member of South Caucasian language family mainly spoken on northeastern coastline of Turkey. It constitutes the first steps to create a gen- eral computational model for word form recognition and production for Laz by building a rule-based morphological anal- yser using Helsinki Finite-State Toolkit (HFST). The evaluation results show that the analyser has a 64.9% coverage over a corpus collected for this study with 111,365 tokens. We have also performed an error analysis on randomly selected 100 tokens from the corpus which are not covered by the analyser, and these results show that the errors mostly result from Turkish words in the corpus and missing stems in our lexicon. 1 Introduction The Laz language, which is mainly spoken on the northeastern coastline of Turkey and also in some parts of Georgia has been recorded as a ‘defi- nitely endangered1’ language in UNESCO Atlas of the World’s Languages in Danger. It belongs to South Caucasian language family2 with the number of speakers estimated to be between 130,000 and 150,000 according to UNESCO 2001 records and between 250,000 and 500,000 according to more recent studies (Haznedar, 2018). Until the 1920s it was a spoken language with only some written col- lection of Laz grammar and folklore studies. Later, İskender Tzitaşi became the pioneer in developing a 1UNESCO defines the degree of definitely endangered as the situation in which “children no longer learn the language as mother tongue in the home”. 2The Southwest Caucasian language family consists of four languages: Svans, Mingrelians, Georgian and Laz. writing system for Laz based on Latin alphabet and later ‘Lazuri Alboni’ (Laz Alphabet) and only after 1990s, the written texts started to come out as sev- eral associations were founded for the preservation of Laz language and culture. Now with all these efforts, Laz has been thought in public schools in Turkey as an elective language course since 2013 (Kavaklı, 2015; Haznedar, 2018). There is not much research on lexicon and syntax of Laz and the first academic level research studies began by the end of 20th century. In 1999, the first dictionary for Laz (Turkish—Laz) was prepared and published by İsmail Bucaklişi and Hasan Uzun- hasanoğlu. The following years, Bucaklişi also pub- lished the first Laz grammar book (Kavaklı, 2015) and has begun teaching Laz at Boğaziçi University in İstanbul as an elective course since 2011. The foundation of the Lazika Publishing Collective in 2011 has given rise to the publication of more than 70 books on Laz language and literature (Kavaklı, 2015). Laz language is only one of many that faces the danger of extinction. By the end of this century, many will …

Content truncated to 500 words