Developing Statistical Methods for Measures of Word Difficulty for Diverse Readers

2017 Fiscal Year Final Research Report

• Project/Area Number: 15K16059
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Intelligent informatics
• Research Institution: National Institute of Advanced Industrial Science and Technology (2016-2017); Tokyo Metropolitan University (2015)
• Principal Investigator: Ehara Yo 国立研究開発法人産業技術総合研究所, 情報・人間工学領域, 研究員 (60738029)
• Project Period (FY): 2015-04-01 – 2018-03-31
• Keywords: 語学学習支援 / 語彙学習 / 項目反応理論 / 機械学習 / 単語難易度 / 能力測定 / テスト理論 / e-learning

Outline of Final Research Achievements

In this study, we developed a statistical method for the construction of word-difficulty measures. Existing measures require intensive effort to build and do not　account for the various specialties of language learners. Nor are they comparable between languages. The method we constructed can estimate the probability that a second-language learner knows a word by taking their specializations into account, using multiple language resources. This method could help estimate the psychological properties of words, such as word familiarity and word imagery. We also constructed datasets for measures of word difficulty, as well as a method for estimating the translation ability of language learners by using only vocabulary tests.

Free Research Field

自然言語処理，知能情報学