TEXT ANALYSIS BASED WORD COMPLETION SYSTEM FOR F KEYBOARD

Öz

Word completion is an assistive technology tool that aims to reduce the number of keys that the user must press. In early studies, it generally handled as an augmentative and alternative communication tool. However, recently word completion systems have also gained importance in different fields such as mobile devices, machine translation and search engines. In this study, a word completion system based on text analysis was developed for F keyboard, Standard Turkish Keyboard. 160 texts of clerk recruitment exams performed by the Ministry of Justice were used to test the developed system due to the intensive keyboard use of the court–house employees. Words’ unigram and bigram frequencies of these texts were extracted and stored in a database. By using this database, the system presents a prediction list of 8 alternative words to the user during the writing process. With the help of defined key combinations, the user can select a word from the prediction list and the selected word is completed automatically by the system. The performance of the system was evaluated in terms of keystroke saving. 50.98% of average keystroke saving rate was provided at the test process done with 15 randomly selected texts among the all texts. 

Anahtar Kelimeler

• Alternative Technology,Word Completion,Word Prediction,Keystroke Saving,F Keyboard

F KLAVYE İÇİN METİN ANALİZİ TABANLI KELİME TAMAMLAMA SİSTEMİ

Öz

Kelime tamamlama, kullanıcının basması gereken tuş sayısını azaltmayı hedefleyen bir yardımcı teknoloji aracıdır. İlk çalışmalarda, genellikle bir alternatif ve destekleyici iletişim aracı olarak ele alınmıştır. Fakat son dönemlerde kelime tamamlama sistemleri; mobil cihazlar, makine çevirisi ve arama motorları gibi farklı alanlarda da önem kazanmıştır. Bu çalışmada, Standart Türkçe Klavye olan F klavye için metin analizi tabanlı bir kelime tamamlama sistemi geliştirilmiştir. Geliştirilen sistemi test etmek için zabıt kâtiplerinin yoğun klavye kullanımından dolayı Adalet Bakanlığı tarafından yapılan zabıt kâtipliği sınavlarına ait 160 metin kullanılmıştır. Bu metinlerdeki kelimelerin unigram ve bigram frekansları çıkarılmış ve bir veritabanında tutulmuştur. Sistem, bu veritabanını kullanarak yazma işlemi esnasında kullanıcıya 8 adet alternatif kelimeden oluşan bir tahmin listesi sunmaktadır. Tanımlanan tuş kombinasyonları yardımıyla kullanıcı tahmin listesinden bir kelime seçebilmekte ve seçilen kelime sistem tarafından otomatik olarak tamamlanmaktadır. Geliştirilen sistemin performansı tuşlama tasarrufu açısından değerlendirilmiştir. Tüm metinler arasından rastgele seçilen 15 metin ile yapılan test işleminde ortalama %50,98 tuşlama tasarrufu sağlanmıştır. 

Anahtar Kelimeler

• Alternatif Teknoloji,Kelime Tamamlama,Kelime Tahmin,Tuşlama Tasarrufu,F Klavye

Kaynakça

1. AB, 2014. Adalet Bakanlığı, Zabıt Kâtipliği Uygulama Sınavı Metinlerine İlişkin Duyuru, http://www.pgm.adalet.gov.tr/duyuru/2014/eylul/zbt.htm (Erişim Tarihi: 20.07.2015)
2. AB, 2015. Adalet Bakanlığı, Zabıt Kâtipliği Uygulama Sınavı Metinlerine İlişkin Duyuru http://www.pgm.adalet.gov.tr/duyuru/2015/subat/metin.htm (Erişim Tarihi: 20.07.2015)
3. Al-Mubaid, H., 2003. Context-based word prediction and classification. In: Proceedings of the 18th International Conference on Computers and their Applications CATA'2003, Hawaii-USA, 384-388.
4. Al-Mubaid, H., 2007. A Learning-Classification Based Approach for Word Prediction. Int. Arab J. Inf. Technol, 4, 264-271.
5. Anson, D., Moist, P., Przywara, M., Wells, H., Saylor, H. and Maxime, H., 2006. The effects of word completion and word prediction on typing rates using on-screen keyboards. Assistive technology, 18, 146-154.
6. Bower, J., Furuuchi, K., Liu, S., Morimoto, K., Robbins, D., Laughlin, C. and Davis, P., 2011. Context based word prediction. U.S. Patent No. 7,912,700.
7. Chronqvist, F. A., 2010. Electronic device with text prediction function and method. U.S. Patent No. 12/363,799.
8. Darragh, J. J., Witten, I. H. and James, M. L., 1990. The reactive keyboard: A predictive typing aid. Computer, 23, 41-49.

9. Fazly, A., 2002. The use of syntax in word completion utilities. MSc Thesis, University of Toronto, Canada.
10. Fazly, A. and Graeme, H., 2003. Testing the efficacy of part-of-speech information in word completion. In: Proceedings of the 2003 EACL Workshop on Language Modelling for Text Entry Methods. Association for Computational Linguistics, Budapest-Hungary, 9-16.
11. Foster, G., Isabelle, P. and Plamondon, P., 1996. Word completion: A first step toward target-text mediated IMT. In: Proceedings of the 16th conference on Computational Linguistics- Association for Computational Linguistics., 1, 394-399.
12. Garay-Vitoria, N. and Gonzalez-Abascal, J., 1997. Intelligent word-prediction to enhance text input rate (a syntactic analysis-based word-prediction aid for people with severe motor and speech disability). In: Proceedings of the 2nd international conference on intelligent user interfaces, New York-USA, 241-244.
13. Handley-More, D., Deitz, J., Billingsley, F. F., & Coggins, T. E., 2003. Facilitating written work using computer word processing and word prediction. American Journal of Occupational Therapy. 57 (2): 139-151.
14. HaCohen-Kerner, Y. and Greenfield, I., 2012. Basic word completion and prediction for Hebrew. In String Processing and Information Retrieval, Springer Berlin Heidelberg, 237-244.
15. Hunnicutt, S., and Carlberger, J., 2001. Improving word prediction using Markov models and heuristic methods. Augmentative and Alternative Communication, 17, 255-264.
16. Karabulut, B., 2016. F Klavye için Metin Analizi Tabanlı Kelime Tamamlama Sistemi Tasarımı. MSc Thesis, University of Kırıkkale, Turkey.
17. Karunarathne, M. S., Nanayakkara, L. D. J. F. and Ponnamperuma, K., 2013. Sentence prediction on SMS in Sinhala language. International Journal of Scientific and Research Publications, 3, 392.
18. Li, J. and Graeme, H., 2005. Semantic knowledge in word completion. In: Proceedings of the 7th international ACM SIGACCESS conference on Computers and accessibility, ACM, Baltimore-USA, 121-128.
19. Miller, J. W., 1998. Word prediction system. U.S. Patent No. 5,805,911.
20. MacKenzie, I. S. and Zhang, X., 2008. Eye typing using word and letter prediction and a fixation algorithm. In: Proceedings of the 2008 symposium on Eye tracking research & applications, Savannah-Georgia, 55-58.
21. MEB, 2012. Millî Eğitim Bakanlığı, Hızlı Yazma, http://kozanhem.meb.k12.tr/meb_iys_dosyalar/01/11/113116/dosyalar/2012_12/05104825_hzlyazadaletmegep.pdf (Erişim Tarihi: 06. 09. 2015)
22. Miller, T. and Wolf, E., 2006. Word completion with latent semantic analysis. In Pattern Recognition, 2006. ICPR 2006. 18th International Conference on Pattern Recognition, IEEE, Hong Kong-China, 1252-1255.
23. Palazuelos, S. E., Aguilera, S., Rodrigo, J. and Godino-Llorente, J. I., 1998. Grammatical and statistical word prediction system for Spanish integrated in an aid for people with disabilities. In ICSLP, Sydney-Australia.
24. Stoop, W., van den Bosch, A. P. J., 2014. Using idiolects and sociolects to improve word prediction. Proceedings of the 14th Conference of the European Chapter of the Association for Computational Linguistics, Gothenburg-Sweden, 318-327.
25. Tam, C., and Wells, D., 2009. Evaluating the benefits of displaying word prediction lists on a personal digital assistant at the keyboard level. Assistive technology, 21, 105-114.
26. Tocci, M., 2012. Predictive word completion, U.S. Patent No. 13/162,319.
27. Thorn, O. K., 2010. Word prediction. U.S. Patent No. 7,698,326.
28. Trnka, K., Yarrington, D., McCoy, K. and Pennington, C., 2006. Topic modelling in fringe word prediction for AAC. In: Proceedings of the 11th international conference on intelligent user interfaces, Sydney-Australia, 276-278.
29. Trnka, K., 2010. Word prediction techniques for user adaptation and sparse data mitigation. PhD Thesis, University of Delaware, USA.
30. Unruh, E., 2010. Text entry with word prediction, completion, or correction supplemented by search of shared corpus. U.S. Patent No. 12/943,856.
31. Van den Bosch, A. and Bogers, T., 2008. Efficient context-sensitive word completion for mobile devices. In: Proceedings of the 10th international conference on Human computer interaction with mobile devices and services ACM, 465-470.
32. Wandmacher, T., & Antoine, J. Y., 2008. Methods to integrate a language model with semantic information for a word prediction component. arXiv preprint arXiv:0801.4716.