• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to previous page

Study on Automatic Review of Essay Based on Generative Adversarial Networks

Research Project

Project/Area Number 18K13240
Research Category

Grant-in-Aid for Early-Career Scientists

Allocation TypeMulti-year Fund
Review Section Basic Section 09070:Educational technology-related
Research InstitutionTohoku Fukushi University

Principal Investigator

Iwata Kazuki  東北福祉大学, 総合マネジメント学部, 講師 (20515457)

Project Period (FY) 2018-04-01 – 2021-03-31
Project Status Completed (Fiscal Year 2020)
Budget Amount *help
¥4,160,000 (Direct Cost: ¥3,200,000、Indirect Cost: ¥960,000)
Fiscal Year 2020: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2019: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2018: ¥2,210,000 (Direct Cost: ¥1,700,000、Indirect Cost: ¥510,000)
Keywords機械学習 / 自然言語処理 / 敵対生成ネットワーク / 自動化 / 文書生成 / 自動評価 / 小論文
Outline of Final Research Achievements

In this study, we focused on adversarial generation networks, one of the unsupervised learning methods without human intervention, as basic research for the automatic evaluation of essays and reports. In addition, we researched document classification tasks using discriminators obtained from adversarial networks. As a result, we found that the performance of discriminators obtained from adversarial learning can classify documents depends on the generator's performance.
In the future, we will add this knowledge and a new algorithm to achieve even higher accuracy in classification.

Academic Significance and Societal Importance of the Research Achievements

本研究は、近年、小論文やレポートによる評価が多様化される中で、機械学習の手法を通してそれらの自動評価システムの構築を目指したものである。研究期間を通して、教師なし学習である敵対生成ネットワークによる識別器の作成が可能であることを見出すことができた。大量のデータを扱うことが可能な教師なし学習による自動評価の構築可能性はビッグデータの利用によって更に性能を向上させることによって、採点者の負担軽減、ならびに、執筆者の学習機会の増加に繋がるものと考えられる。

Report

(4 results)
  • 2020 Annual Research Report   Final Research Report ( PDF )
  • 2019 Research-status Report
  • 2018 Research-status Report
  • Research Products

    (4 results)

All 2020 2018

All Journal Article (3 results) (of which Peer Reviewed: 3 results,  Open Access: 1 results) Book (1 results)

  • [Journal Article] 敵対生成ネットワークによる文書分類2020

    • Author(s)
      岩田一樹
    • Journal Title

      感性福祉研究所年報

      Volume: 22 Pages: 19-31

    • NAID

      120007038998

    • Related Report
      2020 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] 敵対生成ネットワークによる文書生成2020

    • Author(s)
      岩田一樹
    • Journal Title

      感性福祉研究所年報

      Volume: 21 Pages: 53-64

    • NAID

      120006863408

    • Related Report
      2019 Research-status Report
    • Peer Reviewed
  • [Journal Article] 文書生成におけるネットワーク構造依存性2018

    • Author(s)
      岩田一樹
    • Journal Title

      感性福祉研究所年報

      Volume: 20 Pages: 133-140

    • NAID

      40021937234

    • Related Report
      2018 Research-status Report
    • Peer Reviewed
  • [Book] NEW WAYS OF PROMOTING MENTAL WELL-BEING AND COGNITIVE FUNCTIONS2018

    • Author(s)
      Kazuki Iwata, Makoto Ohuchi & Junichi Urushiyama
    • Total Pages
      109
    • Publisher
      Laurea Publications
    • ISBN
      9789517995085
    • Related Report
      2018 Research-status Report

URL: 

Published: 2018-04-23   Modified: 2022-01-27  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi