Title:
複数言語入力での言語処理装置
Document Type and Number:
Japanese Patent JP4088681
Kind Code:
B2
Abstract:
PROBLEM TO BE SOLVED: To provide a language processing method and language processing unit by inputting a plurality of languages, capable of dispensing with preparation for a large amount of conversion rules, and performing language conversion from a conversion object sentence described in a plurality of natural languages to a converted sentence described in another natural language and/or the same natural language and/or language analysis in a plurality of natural languages. SOLUTION: This language processing method performs processing from a processing object sentence described in a plurality of natural languages to a processed sentence described in another natural language and/or the same natural language. In performing the above processing, this method uses a mechanical learning technique for learning language conversion to the above other natural language and/or the above same natural language and/or language analysis on what sentence or text described in the above same natural language the language is easy to become.
Inventors:
Maki Murata
Hitoshi Isahara
Hitoshi Isahara
Application Number:
JP2001331458A
Publication Date:
May 21, 2008
Filing Date:
October 29, 2001
Export Citation:
Assignee:
National Institute of Information and Communications Technology
International Classes:
G06F17/27
Other References:
内元 清貴 他、翻訳メモリとコーパスを用いた学習に基づく訳語選択、電子情報通信学会技術研究報告、日本、社団法人電子情報通信学会、2001.10.10発行、Vol.101、No.351(NLC2001-41)pp.39-46.
村田 真樹 他、サポートベクトルマシンを用いたテンス・アスペクト・モダリティの日英翻訳、電子情報通信学会技術研究報告、日本、社団法人電子情報通信学会、2001.03.09発行、Vol.100、No.700(NLC2000-78)pp.25-32.
村田 真樹 他、サポートベクトルマシンを用いたテンス・アスペクト・モダリティの日英翻訳、電子情報通信学会技術研究報告、日本、社団法人電子情報通信学会、2001.03.09発行、Vol.100、No.700(NLC2000-78)pp.25-32.
Attorney, Agent or Firm:
Hitoshi Shinbo