A STRUCTURAL GRAMMAR FOR UNSUPERVISED LANNA WORD SEGMENTATION FRAMEWORK
Keywords:
Lanna structural grammar, Lanna word segmentation, unsupervised word segmentation, Lanna word segmentation frameworkAbstract
This paper presents a new unsupervised approach to Lanna word segmentation. In aprevious work, we proposed the Lanna structural grammar for computational linguisticsprocessing. Accuracy evaluation shows that the grammar is feasible, valid, and effective.Additionally, we developed the unsupervised word segmentation algorithm that appliedthe Lanna structural grammar to predict words from character sequences and a search forthe optimal segmentation. The proposed algorithm was evaluated on Lanna phrases froma variety of document types. The experimental results show that the segmentation accuracyis effective and practical. It is concluded that the Lanna structural grammar is appropriateand capable for unsupervised Lanna word segmentation. Consequently, our approach issuitable for computational linguistics processing and has the potential to improve thecoverage of morphological analyzers for the Lanna language without explicit wordboundary markers.
References
Chaijaruwanich, J. (2009). Wannaphim Lanna: Literaturepublished in Lanna Dharma 60 books. 1st ed. ChiangMai University Press, Chiang Mai, Thailand,6,402p.
Goldwater, S. , Griffiths, T. L. , and Johnson, M. (2006).Contextual dependencies in unsupervised wordsegmentation. Proceedings of the 21st InternationalConference on Computational Linguistics; July 17-18, 2006; Sydney, Australia, p. 673–680.
Inchai, S. and Wongthong, A. ( 2018) . Introduction toLanna Dhamma alphabet. Chiangrai: The Art andCulture Office Chiangrai, Chiang Rai RajabhatUniversity. Available from: http:/ / ctc. crru. ac.th/topic/708. Accessed date: May 3, 2018.
Lackmuang, P. (2009). Conceptual analysis of idioms inLanna legends, [ M. Ed. thesis] . Department ofEducation Program in Teaching Thai Language,Graduate School, Chiang Mai University, ChiangMai, Thailand, 470p.
Maneechedta, W. (1994). Enculturation process on LannaDhama alphabets, [M.Ed. thesis] . Department ofNonformal Education, Graduate School, ChiangMai University, Chiang Mai, Thailand, 87p.
Mochihashi, D. , Yamada, T. , and Ueda, N. ( 2009) .Bayesian unsupervised word segmentation withnested Pitman- Yor language modeling.Proceedings of the Joint Conference of the 47thAnnual Meeting of the ACL and the 4th InternationalJoint Conference on Natural Language Processingof the Asian Federation of Natural LanguageProcessing; August 2-7, 2009; Singapore, p. 100-108.
Pansook, A. (2008). Status and role changes of the Lannalanguage, [M. A. thesis] . Department of LannaLanguage and Literature, Graduate School, ChiangMai University, Chiang Mai, Thailand, 188p.
Prongthura, N. ( 1982) . Dhamma scripts of NorthernThailand, [M.A. thesis] . Department of OrientalLanguage, Graduate School, Silpakorn University,Bangkok, Thailand, 313p.
Punjaburi, N. ( 2014) . An analytical study of a LannaKhaosaw, the Setthi Hua Wiang, [M.A. thesis] .Department of Oriental Language, GraduateSchool, Silpakorn University, Bangkok, Thailand,424p.
Somsap, S. and Seresangtakul, P. (2013) . Isarn Dharmaword segmentation. Proceedings of the 2013International Conference on Control, Automationand Information Sciences; November 25-28, 2013;Nha Trang, Vietnam, p. 53-57.
Siriratphiriya, K. (2005) . Lanna language: the northernThai characters and how to read them through thestructure of words. 1st ed. Sukhothai ThammathiratUniversity Press, Nonthaburi, Thailand, 234p.
Spiegler, S. , Golenia, B. , and Flach, P. A. ( 2009) .Unsupervised word decomposition with thePromodes algorithm. Proceedings of the 10th CrosslanguageEvaluation Forum Conference onMultilingual Information Access Evaluation: TextRetrieval Experiments; September 30 - October 02,2009; Corfu, Greece, p. 625-632.
Suradet, N. and Yathongkhum, W. ( 2014) . Lannamorphology grammar approach for computationallinguistics processing. Proceedings of the 18thInternational Computer Science and EngineeringConference; July 30 - August 1, 2014; Khon Kaen,Thailand, p. 80–85.
The Center for the Promotion of Arts and Culture, ChiangMai University. ( 2018) . Lanna Kam Muang.Bangkok: Matichon Public Co. , Ltd. Availablefrom: www. matichonweekly. com/ magazinecolumn/lanna. Accessed date: Apr 27, 2018.
Wang, H. , Zhu, J. , Tang, S. , and Fan, X. (2011) . A newunsupervised approach to word segmentation.Comput. Linguist., 37(3):421-454.
Wimonkasem, K. (2006). Northern Thai Dialect. 2nd ed.Silpakorn University Press, Bangkok, Thailand, 366p.
Zhikov, V., Takamura, H., and Okumuran, M. (2010). Anefficient algorithm for unsupervised wordsegmentation with branching entropy and MDL.Proceedings of the 2010 Conference on EmpiricalMethods in Natural Language Processing; October9-11, 2010; Cambridge, MA, USA, p. 832-842.
