2024-03-28T22:53:31Zhttps://ipsj.ixsq.nii.ac.jp/ej/?action=repository_oaipmhoai:ipsj.ixsq.nii.ac.jp:001074362023-04-27T10:00:04Z01164:04179:07434:07764
日本語の語彙的換言知識の質的評価Qualitative Evaluation of Available Japanese Resources for Lexical Paraphrasingjpn概念・項構造http://id.nii.ac.jp/1001/00107412/Technical Reporthttps://ipsj.ixsq.nii.ac.jp/ej/?action=repository_action_common_download&item_id=107436&item_no=1&attribute_id=1&file_no=1Copyright (c) 2014 by the Institute of Electronics, Information and Communication Engineers This SIG report is only available to those in membership of the SIG.長岡技術科学大学電気系長岡技術科学大学電気系梶原, 智之山本, 和英本稿では,現在利用可能な 9 種類の日本語の語彙的換言知識の質的評価を行った.調査の結果,文脈類似度によって Web から自動的に収集された語対に対して,人手で同義関係を判定した語彙的換言知識の質が最も高いことが明らかとなった.本稿での評価の結果を踏まえて,目的に応じた適切な語彙的換言知識を選択することで,語彙的換言の認識および生成を用いた各種自然言語処理タスクの性能改善が期待できる.In this paper, we evaluated the quality of available Japanese resources for lexical paraphrasing. As a result of our examination, it becomes clear that the quality of resources which have been automatically collected by context similarity from the Web and whose synonymous relations are judged by human hand are the highest. Based on the result of the evaluation in this paper, we can expect the performance improvement of various NLP tasks using lexical paraphrase recognition or generation by choosing appropriate resources for lexical paraphrasing according to the purpose.AN10115061研究報告自然言語処理(NL)2014-NL-2198162014-12-092014-12-04