About FOKS

History

FOKS was originally constructed by Slaven Bilac as part of his PhD research at the Tokyo Institute of Technology. The original version is still available here and continues to be used for mobile phones. The current system was rebuilt by Lars Yencken, a current PhD student, who works on maintaining and extending it.

Features

FOKS is a Japanese-to-English dictionary which allows lookup by reading, and can correct for many common types of reading errors, in particular:

Error type	Description
Choice of reading	A single kanji character can have many possible readings, depending on the word it is in. FOKS still shows the right results if the wrong reading for a kanji is entered. Example: 山車 /だし/ "festival float", searched for as "やまくるま", "やましゃ" or "さんしゃ".
Vowel length	Learners often have trouble distinguishing between long and short vowels, especially if their native language does not make this distinction. Example: 大阪 /おおさか/ "Osaka" searched for as "おさか".
Palatalization	Learners often mistake palatalized sounds with their non-palatalized counterparts, and vice-versa. Example: Confusing /りゅ/ and /る/.

Keitai (Cellular) Phone Interface

FOKS is accessible from a number of mobile phone interfaces, in particular EZWEB (WAP), I-mode or J-Sky access methods. Only phones able to display and input Japanese characters can be used. However, the interfaces have been tested in very limited fashion so there might be some problems depending on the make of your phone. If you find an error please contact foks-admin@csse.unimelb.edu.au with a description of your problem. This interface is not actively being developed, and uses an older version of FOKS.

Following URLs can be used to access the respective interfaces:

Method	Url
EZWEB	http://www.foks.info/keitai/ez/
I-mode	http://www.foks.info/keitai/i/
J-Sky	http://www.foks.info/keitai/j/
Other WAP devices	http://www.foks.info/keitai/wap/

Resources

FOKS is built from a number of freely available resources, and word frequency counts from a non-free corpus.

Name	Description
Edict	Edict is a freely available Japanese-English dictionary, maintained by Jim Breen.
Compdic	A dictionary of computer terms.
Enamdict	A dictionary of proper nouns, such as place and person names.
Kanjidic	A dictionary of kanji readings and meanings.
Japan post gazetteer	A place hierarchy mined from Japanese postal districts.

Related publications

For a more detailed description of FOKS and its error models, please have a look at some of the publications related to FOKS:

• Lars Yencken and Timothy Baldwin. 2005. Efficient grapheme-phoneme alignment for Japanese, in Proceedings of the Australasian Language Technology Workshop 2005, Sydney, Australia, pp. 143-151
• Slaven Bilac. 2005. Automatically extending the dictionary to increase its coverage and accessibility. PhD thesis, Tokyo Institute of Technology.
• Slaven Bilac. 2002. Intelligent Dictionary Interface for Learners of Japanese. Masters's thesis, Tokyo Institute of Technology. (ps, pdf)
• Slaven Bilac, Timothy Baldwin, and Hozumi Tanaka. 2002. "Bringing the Dictionary to the User: The FOKS System". In Proc. of the 19th International Conference on Computational Linguistics (COLING2002), pages 89--95. (ps, pdf)
• Timothy Baldwin, Slaven Bilac, Ryo Okumura, Takenobu Tokunaga, and Hozumi Tanaka. 2002. "Enhanced Japanese electronic dictionary look-up".In Proc. of LREC2002. (ps)
• Ryo Okumura. 2001. Basic research on an intelligent dictionary interface for learners of the Japanese language. Bachelor's thesis, Tokyo Institute of Technology.(In Japanese) (ps)
• Timothy Baldwin and Hozumi Tanaka. 2000. "A comparative study of unsupervised grapheme-phoneme alignment methods". In Proc. of the 22nd Annual Meeting of the Cognitive Science Society (CogSci 2000), pages 597--602, Philadelphia. (ps, pdf)
• Timothy Baldwin and Hozumi Tanaka. 1999. "The applications of unsupervised learning to Japanese grapheme-phoneme alignment".In Proc.of ACL Workshop on Unsupervised Learning in Natural Language Processing, pages 9--16, University of Maryland. (ps, pdf)