Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
Last revision Both sides next revision
nvdajp [2012/09/12 00:14]
nishimotz
nvdajp [2012/09/12 16:03]
nishimotz
Line 1: Line 1:
-~~SLIDESHOW~~ +NVDA workshop slide moved to [[NVDA Japanese ​enhancements]]
- +
-====== ​NVDA Japanese enhancements ====== +
- +
-  * The Localization of NVDA for Japanese Language Users +
-  * Takuya Nishimoto, Director of [[http://​en.sourceforge.jp/​projects/​nvdajp/​|NVDA Japanese Team]] +
-  * nishimotz @ gmail.com / Twitter @nishimotz +
-  * to appear in  +
-    * [[http://​workshop.nvda.jp/​|NVDA Workshop in Japan]] (Sep.16,​2012) +
-    * NVDA Japanese ​public meeting (Sep.17,​2012) +
- +
-===== Prehistory ===== +
- +
-  * 1980s: Japanese screen reader. character description. +
-  * 1996: Japanese screen reader for Windows released, known as '​95Reader'​. +
-  * 1997: IBM Home Page Reader. male voice for text, female voice for link. +
-  * 1998: '​PC-Talker'​ released. best selling in Japan. +
-  * 1999: '​95Reader'​ with Japanese braille support. +
-  * 2001: JAWS for Windows from IBM Japan. +
-  * 2006: Japanese mobile phone which announce the input method. +
- +
-===== History ===== +
- +
-  * 2006: NVDA released. +
-  * 2008: Internet Technology Research Committee (ITRC) Universal Access to the Internet (UAI) started discussions of NVDA Japanese support. (Prof. Takayuki Watanabe, Mitsue-Links Co.,Ltd) +
-  * 2010: Japanese speech engine, input method support. (Takuya Nishimoto, Masataka Shinke) +
-  * 2011: Japanese braille support. 64bit system support. +
-  * 2012: NVDA Japanese team restructured. +
- +
-===== Internationalization ===== +
- +
-  * NVDA Translation team: 40 languages. +
-    * messages +
-    * documents +
-    * symbols, character descriptions +
-    * add-ons: OCR, Vocalizer +
-    * community web site +
- +
-  * Dependency (Japanese support unavailable) +
-    * eSpeak: multi-lingual open-source speech synthesizer +
-    * liblouis: multi-lingual open-source braille translator +
- +
-===== Japanese team (2012) ===== +
- +
-  * Team: non-profit developer community +
-    * 24 persons +
-    * weekly Skype meeting +
-  * Japanese Users List +
-    * 179 persons +
-    * e-mail discussions +
-  * Friends +
-    * text book for support volunteers +
-  * Taiwan-Japan collaboration +
-    * bi-monthly Skype meeting +
- +
-===== Works ===== +
- +
-  * Represent Japanese users +
-    * develop enhancement for Japanese +
-    * 2012.2.1jp: more than 2000 downloads +
-    * define requirement for Japanese support +
-  * Support international users of Japanese language +
-    * add-on speech engine +
-  * Participate in international team +
-    * support East-Asia enhancement work +
-    * migration from localization to internationalization +
- +
-===== Japanese requirements ===== +
- +
-  * Why/how Chinese support and Japanese support are different?​ +
-  * Domestic conventions?​ +
-    * screen readers since 1980s +
-  * Characteristics of Japanese language +
-    * explained in [[http://​shop.oreilly.com/​product/​9780596514471.do|CJKV Information Processing, 2nd Edition]] +
-    * uniqueness of writing system, braille system +
-  * Technical limitations +
-    * keyboard, speech engine +
- +
-===== Input methods ===== +
- +
-  * Correct pronunciation,​ wrong characters +
-    * gohenkan http://​bit.ly/​NXaVI2 +
-    * 'ha i ri so u de su': 'Yes, it is an ideal.'​ / 'It is likely to go into inside.'​ +
- +
-  * Same technology +
-    * Input Method Manager (IMM), Text Service Framework (TSF) +
-  * Difference of writing systems +
-    * Candidate is not single character +
-    * State transition is differenct +
- +
-===== State transitions ===== +
- +
-  * Chinese +
-    * initial state +
-    * editing reading string (on the spot) +
-    * candidate selection (composition window) +
-  * Japanese +
-    * initial state +
-    * editing reading string (on the spot), before translation key pressed +
-    * composition string (on the spot), after translation key pressed +
-    * candidate selection (candidate window pop up), translation key pressed twice +
- +
-===== Half shape/full shape ===== +
- +
-  * History: single byte/double byte +
-  * Unicode: shape does not correspond to byte size +
-  * Input method  +
-    * disabled: half shape only +
-    * enabled: half shape and full shape are selectable +
- +
-  * Discrimination +
-    * not necessary for understanding the meaning +
-    * necessary for writing, typing URLs, filling web forms (especially user authorization),​ software development +
- +
-===== Characters 1 ===== +
- +
-  * Latin (alphabet) +
-    * half shape, full shape +
-  * Number +
-    * half shape roman digit, full shape roman digit, ideograph +
-  * Pronunciation syllable (Kana) +
-    * Hiragana: grammatical words, inflectional endings +
-    * Katakana: foreign words. half shape and full shape +
- +
-===== Characters 2 ===== +
- +
-  * Ideographs +
-    * Kanji in Japanese, Hanzi in Chinese, Hanja in Korean +
-    * Chinese traditional > Japanese > Chinese simplified +
-    * typically many readings for Japanese +
-  * Symbols, punctuations +
-    * 600 full shape symbols by Japanese standard. +
-    * some symbols available for half shape and full shape +
- +
-===== Writing system 1 ===== +
- +
-  * Sentence contains both syllables and ideographs +
-  * Word is not separeted with spaces +
-  * Pronunciation of ideograph depends on the context +
- +
-  * Input method +
-    * syllable-to-ideograph conversion using dictionary +
- +
-===== Writing system 2 ===== +
-  * Braille +
-    * popular six-dot system +
-      * syllable based symbols are defined for Japanese braille. +
-      * special rules for word breaks. +
-    * ideographic braille: six-dot system and eight-dot system +
- +
-  * Speech synthesis +
-    * pronunciation of single character (spelling reading) is undefined. +
- +
-  * Morphological analysis: useful in many places +
-    * based on dictionary, and sometimes statics (machine learning) +
- +
-===== Character descriptions 1 ===== +
- +
-  * Chinese: many examples and explanations +
-    * announce first or announce all +
-  * Japanese: one example or explanation +
-    * written in syllable character +
- +
-  * Usage +
-    * input method candidates +
-    * character review +
-    * text edit: very frequently +
- +
-===== Character descriptions 2 ===== +
- +
-  * Charater description  +
-    * spelling reading (with dictionary) +
-    * ideographic character: necessary for descrimination +
-      * both for speech and braille display +
-    * phonetic reading: help listening of speech +
-      * latin character for English +
-      * syllable character for Japanese katakana and hireagana +
- +
-  * Character attributes +
-    * similarity to announce capital letter +
-      * possible with speech, beep, pitch +
-    * half-shape/​full-shape (Latin, katakana, number, symbol) +
-    * types of syllable character: katakana/​hiragana +
- +
-===== Syllable input ===== +
- +
-  * Transliteration:​ Latin-to-syllable (roman)  +
-    * US key array +
-    * Japanese key array +
-    * braille input (Latin six-dot system) +
- +
-  * Direct: (kana) +
-    * Japanese key array +
-    * braille input (Japanese six-dot system) +
- +
-===== Keyboard 1 ===== +
- +
-  * Half-shape/​Full-shape key +
-    * enable or disable input method +
-    * US key array: ALT-tilde is equivalent +
-  * Non-Conversion key +
-    * input mode change +
-    * do not convert reading string (Hiragana) +
-    * convert reading string from Hiragana to Katakana ​   +
-  * Conversion key +
-    * space key is equivarent in most cases +
-  * Katakana/​Hiragana/​Roman key +
-    * with ALT modifier, syllable input mode change +
- +
-===== Keyboard 2 ===== +
- +
-  * Caps Lock key +
-    * in Japanese array, cannot be used as NVDA modifier key +
- +
-  * Variations +
-    * Non-Conversion equivalents:​ Ctrl-U,​I,​O,​P and F6-F9 +
- +
-  * Preferences +
-    * change mode before typing syllables +
-    * (non) convert after typing syllables +
- +
-===== Japanese speech ===== +
- +
-  * JTalk (based on Open JTalk, MMDAgent) +
-    * Hidden Markov Model (HMM) synthesizer +
-    * Mecab text processor +
-    * ultra-fast speech +
-  * JTalk2 +
-    * automatic language change +
-    * JTalk for '​ja'​ +
-    * eSpeak for others +
- +
-===== Applications ===== +
- +
-  * Input method editors (third parties) +
-    * ATOK from Just System, Google Japanese Input, Baidu IME +
-  * PC-Talker companions +
-    * NetReader may be associated with web pages. +
-  * OpwBE +
-  * Hidemaru editor +
-    * ControllerClient enhancement:​ speakSpelling +
- +
-===== Devices ===== +
- +
-  * KGS braille display +
-    * Japanese company: KGS Corporation +
-    * serial port support requested from users +
-    * vendor library depends on serial port emuration +
- +
-===== Braille translator design ===== +
- +
-  * (1) Pre-process +
-    * separation: Japanese, English, computer +
-  * (2) Morphological analysis (Japanese) +
-    * Mecab already used for JTalk speech engine +
-  * (3) Morphological unit processing (Japanese) +
-    * word break, long vowel marks +
-    * Latin, number, Japanese symbol, space +
-    * number translation +
-  * (4) Word break detemination +
-  * (5) Post-process +
-    * merge Japanese and English/​computer +
-    * foreign mark, cap mark, number mark, braille symbols +
-    * make dot patterns +
- +
-===== Future ===== +
- +
-  * Separate interface language and content language +
-    * speech is already multilingual +
-    * multilingual input methods support +
- +
-  * Costly development in Japanese community +
-    * is our language really special? +
- +
-  * Character/​symbol description +
-    * efficiency: for engineer +
-    * correctness:​ for school+
  
nvdajp.txt · Last modified: 2012/09/12 16:03 by nishimotz
www.chimeric.de Valid CSS Driven by DokuWiki do yourself a favour and use a real browser - get firefox!! Recent changes RSS feed Valid XHTML 1.0