Japanese Input Method with Open Data
No video of the event yet, sorry!
In order to input Japanese from the PC, you need to convert Kanji which does not exist in English-speaking countries. The user inputs Kana characters and converts it into Katakana characters or Kanji. In other words, developers must consider how to accept input characters from Kana characters and how to output the candidate Kanji characters to be converted. We have made progress on Japanese input in Linux, but development is somewhat stagnant. Regarding input methods accepting input characters by users, there are not so many developers and it is difficult to keep up with the new UI easily. In addition, the Kana-Kanji conversion system that outputs the Kanji candidate that you want to convert has become a necessity for a huge Japanese corpus for development, so if you are not a big company specializing in big data, It is getting hard to develop. For openSUSE, iBus is installed as an input method, and Mozc developed by Google as Kanji conversion system is installed by default. However, Mozc is currently only for maintenance, it has never become a new system. In order to overcome this situation, I think that a Kana-Kanji conversion system that utilizes open data is necessary. How should a new Kana - Kanji conversion be made? I will organize the problems in kana-kanji conversion so far and propose a method that incorporates open data.
- 2017 October 21 - 17:15
- 30 min
- Room 2
- openSUSE.Asia Summit 2017
- Linux Desktop