A Constraint Solving Web Service for a Handwritten Japanese Historical Kana Reprint Support System
Reading Japanese historical manuscripts is one of the first steps for researching Japanese classical literature. It is, however, difficult and time-consuming even for Japanese people to read such manuscripts since they are handwritten in cursive style and may contain different characters from those currently used. We formulated the human process for reading Japanese historical manuscripts as a constraint satisfaction problem, and proposed a framework to assist the process. In this paper, we present a constraint solving Web service as a part of a system based on the framework. To realize the Web service, we added a Web service layer to our constraint solver previously implemented in Ruby as a UNIX command. Thanks to the loose coupling realized by the Web service, any programming language can be used for implementation of other parts of the whole system. We experimentally confirmed the solver as a Web service is faster than that as a UNIX command if both the solver and a client are connected to a same local area network. We finally summarized technical issues concerning the system based on the framework.
KeywordsNatural language processing Morphological analysis Constraint solving Web service Reprint Historical document
This work was supported by JSPS KAKENHI Grant Number JP16K00463.
- 2.Watanabe, S., Suzuki, T., Aiba, A.: Reducing of the number of solutions using adjacency relation of words in recognizing historical KANA texts. IPSJ J. 56(3), 951–959 (2015)Google Scholar
- 3.Sando, K., Suzuki, T., Aiba, A.: A constraint solving web service for recognizing historical Japanese KANA texts. In: Rocha, A.P., van den Herik, J. (eds.) Proceedings of the 10th International Conference on Agents and Artificial Intelligence, ICAART 2018, vol. 2, Funchal, Madeira, Portugal, 16–18 January 2018, pp. 257–265. SciTePress (2018)Google Scholar
- 5.Reizei, T.: Tales of Ise (photocopy). Kasama Shoin (1994)Google Scholar
- 6.Yamamoto, S., Osawa, T.: Labor saving for reprinting Japanese rare classical books. J. Inf. Process. Manag. 58(11), 819–827 (2016)Google Scholar
- 7.Terasawa, K., Kawashima, T.: Word spotting online. In: Proceedings of the Computers and the Humanities Symposium, vol. 2011, pp. 329–334 (2011)Google Scholar
- 8.Manmatha, R., Han, C., Riseman, E.M., Croft, W.B.: Indexing handwriting using wordmatching. In: Proceedings of the First ACM International Conference on Digital Libraries, DL 1996, pp. 151–159. ACM, New York (1996)Google Scholar
- 11.Hayasaka, T., Ohno, W., Kato, Y., Yamamoto, K.: Trial production of application software for machine transcription of Hentaigana by deep learning. In: Proceedings of the 31st Annual Conference of the Japanese Society for Artificial Intelligence (2017)Google Scholar
- 12.Hayasaka, T., Ohno, W., Kato, Y., Yamamoto, K.: Recognition of kuzushiji (hentaigana and cursive script) by deep learning (ver.0.5.1) (2017). http://vpac.toyota-ct.ac.jp/kuzushiji/
- 13.Yamada, S., Shibayama, M.: An estimation method of unreadable historical character for manuscripts in fixed forms using n-gram and OCR. IPSJ SIG Notes 2003(59), pp. 17–24, May 2003Google Scholar
- 18.Ogiso, T., Komachi, M., Den, Y., Matsumoto, Y.: UniDic for early middle Japanese: a dictionary for morphological analysis of classical Japanese. In: Calzolari, N., et al. (eds.) Proceedings of the Eight International Conference on Language Resources and Evaluation (LREC 2012), Istanbul, Turkey. European Language Resources Association (ELRA), May 2012Google Scholar