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

An Development of automated short-answer scoring system based on deep learning without using supervised scoring data

Research Project

Project/Area Number 20H04300
Research Category

Grant-in-Aid for Scientific Research (B)

Allocation TypeSingle-year Grants
Section一般
Review Section Basic Section 62030:Learning support system-related
Research InstitutionThe National Center for University Entrance Examinations

Principal Investigator

Ishioka Tsunenori  独立行政法人大学入試センター, 研究開発部, 教授 (80311166)

Co-Investigator(Kenkyū-buntansha) 中川 正樹  東京農工大学, 学内共同利用施設等, 特任教授 (10126295)
峯 恒憲  九州大学, システム情報科学研究院, 准教授 (30243851)
須鎗 弘樹  千葉大学, 大学院工学研究院, 教授 (70246685)
宮澤 芳光  独立行政法人大学入試センター, 研究開発部, 准教授 (70726166)
Project Period (FY) 2020-04-01 – 2023-03-31
Project Status Completed (Fiscal Year 2023)
Budget Amount *help
¥16,120,000 (Direct Cost: ¥12,400,000、Indirect Cost: ¥3,720,000)
Fiscal Year 2022: ¥5,460,000 (Direct Cost: ¥4,200,000、Indirect Cost: ¥1,260,000)
Fiscal Year 2021: ¥5,200,000 (Direct Cost: ¥4,000,000、Indirect Cost: ¥1,200,000)
Fiscal Year 2020: ¥5,460,000 (Direct Cost: ¥4,200,000、Indirect Cost: ¥1,260,000)
Keywords自然言語処理 / 自動採点 / 機械学習 / 深層学習 / トランスフォーマー / 手書き文字認識 / アンサンブル学習
Outline of Research at the Start

センター試験など大学入試試験レベルの短答式記述試験の自動採点および人間による採点を支援する実用可能なシステムを試作・実装する。採点は設問ごとに作題者が用意した「模範解答」と「採点基準」に従いシステムがある程度の精度をもった採点計算(自動採点)を行うことを基本とし、その結果を人間が確認・修正できるものとする。このシステムの最大の特徴は「(予め用意された)模範解答」と「(被験者の実際の)記述解答」との意味的同一性や含意性の判定に採点済みの教師データを使わないことにある。予め別に用意された新聞や教科書、Wikipediaなど別のコーパスなどから自動構築した言語モデルによって判定を行う。

Outline of Final Research Achievements

In recent years, research into deep learning methods called recurrent neural networks, especially transformers such as BART, has progressed, and their excellent performance has been proven. Here, we consider written sentences in natural language as time-series data with an order, and process this as input data. We attempted to process written response data from 120K common test trial surveys conducted in 2017 and 2017, from character recognition to automatic scoring using Bart, all at once. Our collaborative research group achieved an average agreement rate of 96% and a minimum of 93% in real-world operations without the manual training wheels used in conventional scoring systems. Additionally, by using a huge amount of data containing 60K questions for each question, we gained new knowledge about the sample size required for deep learning.

Academic Significance and Societal Importance of the Research Achievements

いままでの研究では学習データに用いるサンプルはせいぜい2千件程度であり、どの程度のサンプルがあれば十分な予測ができるかの目安は与えられていなかった。さらに九大グループでは意味的埋め込みと呼ばれる異なったアプローチによる方法を試みた。これら結果については本科研で3件の学会表彰(日本計算機統計学会第35回大会, 学生研究発表賞;Duolingo Award for IMPS 2021;SMASH22 Winter Symposium,準優秀賞)を受け、その成果については日本教育新聞や日経新聞教育面に大きく掲載された。その後、教育工学のトップ国際会議AIED 2022でも論文採択された。

Report

(4 results)
  • 2023 Final Research Report ( PDF )
  • 2022 Annual Research Report
  • 2021 Annual Research Report
  • 2020 Annual Research Report
  • Research Products

    (17 results)

All 2023 2022 2021 2020

All Journal Article (3 results) (of which Int'l Joint Research: 3 results,  Peer Reviewed: 3 results) Presentation (13 results) (of which Invited: 1 results) Book (1 results)

  • [Journal Article] Optimizing Answer Representation using Metric Learning for Efficient Short Answer Scoring2023

    • Author(s)
      Bo Wang, Billy Dawton, Tsunenori Ishioka and Tsunenori Mine
    • Journal Title

      The Pacific Rim International Conference on Artificial Intelligence (PRICAI)

      Volume: 20th Pages: 236-248

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed / Int'l Joint Research
  • [Journal Article] Fully automated short answer scoring of the trial tests for common entrance examinations for japanese university2022

    • Author(s)
      Oka, H., Nguyen, H. T., Nguyen, C. T., Nakagawa, M. & Ishioka, T.
    • Journal Title

      AIED 2022

      Volume: LNCS 13355 Pages: 180-192

    • DOI

      10.1007/978-3-031-11644-5_15

    • ISBN
      9783031116438, 9783031116445
    • Related Report
      2022 Annual Research Report
    • Peer Reviewed / Int'l Joint Research
  • [Journal Article] Handwriting recognition and automatic scoring for descriptive answers in japanese language tests, International Conference on Frontiers in Handwriting Recognition2022

    • Author(s)
      Nguyen, H.T., Nguyen, C. T., Oka, H., Ishioka, T. & Nakagawa, M.
    • Journal Title

      International Conference on Frontiers in Handwriting Recognition, ICFHR 2022

      Volume: LNCS 13639 Pages: 274-284

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed / Int'l Joint Research
  • [Presentation] Automated Short Answer Grading with Rublic-based Semantic Embedding Optimization2022

    • Author(s)
      Wang,B., Ishioka,T., Mine,T.
    • Organizer
      SMASH22 Winter Symposium
    • Related Report
      2022 Annual Research Report 2021 Annual Research Report 2020 Annual Research Report
  • [Presentation] Fully automatic scoring of handwritten descriptive answers in Japanese language tests2022

    • Author(s)
      Nguyen, H.T., Nguyen, C. T., Oka, H., Ishioka, T. & Nakagawa, M.
    • Organizer
      IEICE technical report, PRMU2021-32, 45-50
    • Related Report
      2022 Annual Research Report
  • [Presentation] Visual constraints for generating multi-domain offline handwritten mathematical expressions2022

    • Author(s)
      Ung, H. Q., Nguyen, C. T., Oka, H., Ishioka, T. & Nakagawa, M.
    • Organizer
      IEICE technical report, PRMU2021-69, 54-59
    • Related Report
      2022 Annual Research Report
  • [Presentation] 大学入学共通テスト試行調査における短答式記述答案の完全自動採点2022

    • Author(s)
      岡知樹,N.T.Hung, N.TCuong, 中川正樹, 石岡恒憲
    • Organizer
      言語処理学会第28回年次大会, E3-5, 若手奨励賞
    • Related Report
      2021 Annual Research Report
  • [Presentation] Fully automatic scoring of handwritten descriptive answers in Japanese language tests2022

    • Author(s)
      Hung Tuan Nguyen・Cuong Tuan Nguyen(TUAT)・Haruki Oka(UTokyo)・Tsunenori Ishioka(The National Center for University Entrance Examinations)・Masaki Nakagawa(TUAT)
    • Organizer
      電子情報通信学会 研究会PRMU2021-32, 45-50.
    • Related Report
      2021 Annual Research Report
  • [Presentation] 大学入学共通テスト試行調査における短答式記述答案の完全自動採点2022

    • Author(s)
      岡知樹,N.T.Hung, N.TCuong, 中川正樹, 石岡恒憲
    • Organizer
      言語処理学会第28回年次大会E3-5
    • Related Report
      2020 Annual Research Report
  • [Presentation] 大学入学共通テスト試行調査における記述式問題の自動採点2021

    • Author(s)
      岡知樹,N.T.Hung, N.TCuong, 中川正樹, 石岡恒憲
    • Organizer
      日本計算機統計学会第35回大会, 学生研究発表賞
    • Related Report
      2021 Annual Research Report
  • [Presentation] Short answer scoring of the trial test for Japanese Common University Entrance Examination,2021

    • Author(s)
      Oka,H., Hung,N.T., Cuong,N.T., Nakagawa,M., Ishioka,T.
    • Organizer
      IMPS, Duolingo Award
    • Related Report
      2021 Annual Research Report
  • [Presentation] 共通テストの試行調査国語記述解答データを用いた自動採点のアルゴリズムとその評価2021

    • Author(s)
      石岡恒憲, 岡知樹, N.T.Hung, N.TCuong, 中川正樹
    • Organizer
      日本テスト学会第19回大会発表論文抄録集, 124-125.
    • Related Report
      2021 Annual Research Report
  • [Presentation] 大学入学共通テスト試行調査における記述式問題の自動採点2021

    • Author(s)
      岡知樹,N.T.Hung, N.TCuong, 中川正樹, 石岡恒憲
    • Organizer
      日本計算機統計学会第35回大会
    • Related Report
      2020 Annual Research Report
  • [Presentation] 共通テストの試行調査国語記述解答データを用いた自動採点のアルゴリズムとその評価2021

    • Author(s)
      石岡恒憲, 岡知樹, N.T.Hung, N.TCuong, 中川正樹
    • Organizer
      日本テスト学会第19回大会発表論文抄録集, pp.124-125
    • Related Report
      2020 Annual Research Report
  • [Presentation] 短答式試験における自動採点のための概念辞書を用いたデータ拡張手法の提案2021

    • Author(s)
      加藤博之・石岡恒憲・峯恒憲
    • Organizer
      信学技報, vol. 120, no. 344, AI2020-15, pp. 7-12
    • Related Report
      2020 Annual Research Report
  • [Presentation] AI-based+ Automated Short-answer Scoring System2020

    • Author(s)
      Ishioka, T.
    • Organizer
      Digital World 2020
    • Related Report
      2020 Annual Research Report
    • Invited
  • [Book] 自動採点研究のこれから.「英語教育研究における自動採点 現状と課題」2020

    • Author(s)
      石岡恒憲 (石井雄隆・近藤悠介(編))
    • Total Pages
      157
    • Publisher
      ひつじ書房
    • ISBN
      9784823410604
    • Related Report
      2020 Annual Research Report

URL: 

Published: 2020-04-28   Modified: 2025-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi