Updated on 2022/05/17

写真a

 
KUBO, Kei
 
Affiliation
Affiliated organization, Center for Japanese Language
Job title
Associate Professor(non-tenure-track)

Education

  • 2010.04
    -
    2013.03

    Kyoto University   Graduate School of Human and Environmental Studies  

  • 2008.04
    -
    2010.03

    Kyoto University   Graduate School of Human and Environmental Studies  

  • 2003.04
    -
    2007.03

    Kobe City University of Foreign Studies   Faculty of Foreign Studies   Department of English Studies Evening Course  

Degree

  • 京都大学   博士

Research Experience

  • 2021.09
    -
    Now

    Waseda University   Center for Japanese Language

  • 2019.04
    -
    2021.09

    Ritsumeikan University   Ritsumeikan International

  • 2015.04
    -
    2021.09

    Ryukoku University

  • 2016.04
    -
    2019.03

    Kyoto Notre Dame University

  • 2015.04
    -
    2019.03

    Kyoto University

  • 2013.04
    -
    2019.03

    Osaka University   Center for Japanese Language and Culture

  • 2017.10
    -
    2018.03

    Ryukoku University   Faculty of Science and Technology

▼display all

Professional Memberships

  • 2015.04
    -
    Now

    日本語教育学会

  • 2011.04
    -
    Now

    言語処理学会

  • 2011
    -
    Now

    日本語文法学会

  • 2008.04
    -
    Now

    日本認知言語学会

 

Research Areas

  • Japanese linguistics

  • Japanese language education

Research Interests

  • 日本語学

  • 認知言語学

  • 否定

  • 接辞表現

  • 日本語教育

Papers

  • Development of Language Ability Measurement System (KOTOBAKARI) using Voice Recognition

    Miyabe Mai, Shikata Shuko, Kubo Kay, Aramaki Eiji

    Journal of Natural Language Processing   25 ( 1 ) 33 - 56  2018

     View Summary

    <p>There is a growing need for automatic measurements of language ability. With increase in the aging population, the number of older adults with dementia is expected to increase. Language disorder is considered one of the most fundamental symptom with which dementia can be detected. The identification of language defects particular to dementia symptoms may contribute to the early detection of dementia. Currently, many foreign students learn Japanese at educational institutions, which are required to provide appropriate evaluations to their students. However, differences occur because evaluators judge the language abilities of foreign students subjectively. The measured results from an automatic language ability measurement system could be used for objective evaluations. Although traditional studies manually measured language abilities, such methods are often costly. In this study, we propose the automatic language ability measurement system, "KOTOBAKARI," as an alternative to traditional manual measurement methods. This system is expected to reduce the cost and time of measuring the language ability using the following steps: (1) voice recognition to obtain text data and (2) quantitative indicators for automated language ability measurement. We verify the capacity of voice recognition to measure language ability by comparing language ability to a threshold value and by assessing language ability score changes. The experimental results show that it is difficult to judge language ability by threshold value comparison, using our system. On the contrary, the language ability score of some individuals, measured with our system, was correlated with the correct score. This result indicates that these individuals can use our system to assess language ability changes by continuous measurements. </p>

    DOI CiNii

  • 日本語接辞にみられる否定の意味的多様性とその体系的分類

    久保 圭

       2017.03

    Authorship:Lead author

  • 言秤 認知症スクリーニング支援のための言語能力自動測定システムの開発

    宮部 真衣, 四方 朱子, 久保 圭, 荒牧 英治

    医療情報学連合大会論文集   35回   1296 - 1299  2015.11

  • Collection of Modal Records with Location Information via Citywalk

    IPSJ Journal   56 ( 1 ) 207 - 218  2015.01

     View Summary

    A concept of "modality" has been proposed in architecture field. "Modality" means the totality of what we are aware of. Information about modality is useful in various situations, such as sightseeing, daily life, and city plan. In order to collect modality-records using information-processing technology, we held the event "100ninmap" that people walk about town. The results revealed three findings: (1) Almost half of collected modal records are location-dependent records. (2) In order to collect modality-records efficiently, the event takes a little ingenuity that keeps someone from being bored. (3) Through the event, we can collect not only information about shopping, eating and drinking but that about daily life, entertainments, and so on.

    CiNii

  • 音声認識による認知症・発達障害スクリーニングは可能か?-言語能力測定システム"言秤"の提案-

    宮部真衣, 四方朱子, 久保圭, 荒牧英治

    ワークショップ2014 (GN Workshop 2014) 論文集   ( 2014 ) 1 - 8  2014.11

    CiNii

  • Another Medical Big Data : Turning Illness Narrative to Practical Use

    ARAMAKI Eiji, SHIMAMOTO Yumiko, KUBO Kay, NAKAMURA Tetsuaki, SHIKATA Shuko, MIYABE Mai, Eiji Aramaki, Yumiko Shimamoto, Kay Kubo, Tetsuaki Nakamura, Shuko Shikata, Mai Miyabe

    Journal of the Japanese Society for Artificial Intelligence   29 ( 6 ) 599 - 603  2014.11

    DOI CiNii

  • 類義語を共起関係から紐解く : コーパス分析を用いた指導の有用性

    久保 圭, クボ ケイ

    大阪大学日本語日本文化教育センター授業研究   ( 12 ) 23 - 31  2014.03

     View Summary

    日本語には、似た意味をもつさまざまな言葉がある。たとえば、「誤りを直すJという意味をもつ日本語のサ変名詞には、修正、是正、訂正、改鼠などがある。しかし、その意味的なニュアンスや用法(どのような対象における誤りを直すのか、など)は一律ではなく、すべて異なっていると考えられる。これは、上記に挙げた類義語がすべて同じ意味をもつのであれば、これほど多様な表現は必要ないと考えられるためである。上記のような類義語の使い分けの習得に関心を寄せる留学生は少なくないように感じられる。しかし、日本語の言語直感をもたない彼らにとって、その学習は困難をともなうものであるため、類似した意味の言葉を教授する際には、効果的な説明を準備しなくてはならないだろう。本論文では、日本語学習者に向けた教育の現場において、類似した意味をもついくつかの語の意味と使い分けを説明する際に、大規模コーパスから得られる、各類義語の共起関係に関する調査データの活用が有効であることを示唆する。

    DOI CiNii

  • Defining patients with depressive disorder by using textual information

    Tetsuaki Nakamura, Kay Kubo, Yasuyuki Usuda, Eiji Aramaki

    AAAI Spring Symposium - Technical Report   SS-14-01   39 - 44  2014

     View Summary

    Depression has been one of the serious social problems of the modern society, due to the prolonged economic stagnation, the increasing number of unemployment rate, and so on. Diseases that require long-term treatments, such as cancer and chronic diseases, are also the major cause of depression, and, eventually, that often leads the patients to committing suicide. Such circumstance requires some urgent means to prevent depression, especially of those who are suffering diseases. In this study, to propose one of the tools for preventing depression or to detect depression at its early stage of suffering, we analyze 200 blog articles to set-up an "evaluation index" that is capable of measuring the tendency of depression from written texts of the long-term patients. Firstly, we collect and analyze blog articles written by the patients who are suffering depression, as well as the articles written by those who are not suffering depression as a comparison. From the analysis, we propose an index for depression measurement. With this index, secondly, we make an analysis on the articles that are posted on a special SNS, which is particularly available for long-term patients. It comes out that posting articles on SNS for more than 12 months may mitigate depression (or depressive symptoms). This outcome suggests the possibility of SNS posting against depression for the long-term patients. Copyright © 2014, Association for the Advancement of Artificial Intelligence. All rights reserved.

  • 街歩きで作り出す都市の様相地図―位置情報付きの様相記録収集の取り組み―

    宮部真衣, 北雄介, 久保圭, 荒牧英治

    ワークショップ2013 (GN Workshop 2013) 論文集   ( 2013 ) 1 - 8  2013.11

    CiNii

  • 日本語の否定接頭辞に関する認知言語学的分析

    久保 圭

       2010.03

    Authorship:Lead author

▼display all

Books and Other Publications

  • 文章を科学する

    李, 在鎬, 石黒, 圭, 伊集院, 郁子, 河原, 大輔, 久保, 圭, 小林, 雄一郎, 長谷部, 陽一郎, 樋口, 耕一( Part: Joint author, 第7章(pp. 118-137))

    ひつじ書房  2017.10 ISBN: 9784894768819

Misc

  • Corpus and Education of Japanese as the Foreign Language. Series of the Japanese Corpus, vol. 5. Tokyo: Asakura Shoten.

    Kubo Kei

    Mathematical Linguistics   31 ( 5 ) 370 - 372  2018

    Book review, literature introduction, etc.  

    DOI CiNii

  • 日本語を測る -言語能力測定システムの日本語教育への応用可能性について-

    久保 圭, 宮部 真衣, 四方 朱子, 李 在鎬, 荒牧 英治

    第77回全国大会講演論文集   2015 ( 1 ) 23 - 24  2015.03

     View Summary

    現在、本邦では13万人を超える外国人留学生が日本語を学んでいる。各日本語教育機関には、留学生の日本語熟達度に対して適切に評価を与えることが求められているが、評価の難易度はその対象によって異なる。選択問題など、問いに対する解が明示的に設けられていることの多い読解や聴解の試験については、その評価が比較的容易であるが、作文課題やスピーチ・テストについては、教師によって評価基準に主観が入る可能性が多分にあり、客観的な評価が非常に困難であるといえる。そこで、本発表では言語能力(語彙量・語彙の特殊性・語彙の難しさなど)を数量的に測定するシステムを紹介し、日本語教育への応用可能性について探る。

    CiNii

  • Aging and Language : Blog Texts as an Aging Index

    ARAMAKI Eiji, KUBO Kay, SHIKATA Shuko

    IEICE technical report. Data engineering   114 ( 173 ) 131 - 136  2014.08

     View Summary

    Preceding study claims that one's language abilities develop over long period of time and improve with age. On the other hand, some study reports that some parts of language abilities, such as grammatical ability, show some decrease in elder people. Since one's language ability is often shown as the aggregation of multiple human abilities, it is difficult to solely extract his/her language ability out of his/her written texts. This study, thus, analyzes texts by using multiple linguistic measures. The corpora cover school students (children attending primary to junior high school, age 6 to 15 years old), elders (age 50 to over 80 years old), Japanese as the Second Language learners, and a dementia patient (Alzheimer type). As a result, this study shows that the lexical richness decreases, and difficult vocabularies tend to be especially lost from elders. This study also displays the possibility of detecting dementia in its early stage.

    CiNii

  • Aging and Language: Blog Texts as an Aging Index

    Eiji Aramaki, Kay Kubo, Shuko Shikata

    IPSJ SIG Notes   2014 ( 23 ) 1 - 6  2014.07

     View Summary

    Preceding study claims that one's language abilities develop over long period of time and improve with age. On the other hand, some study reports that some parts of language abilities, such as grammatical ability, show some decrease in elder people. Since one's language ability is often shown as the aggregation of multiple human abilities, it is difficult to solely extract his/her language ability out of his/her written texts. This study, thus, analyzes texts by using multiple linguistic measures. The corpora cover school students (children attending primary to junior high school, age 6 to 15 years old), elders (age 50 to over 80 years old), Japanese as the Second Language learners, and a dementia patient (Alzheimer type). As a result, this study shows that the lexical richness decreases, and difficult vocabularies tend to be especially lost from elders. This study also displays the possibility of detecting dementia in its early stage.

    CiNii

  • 言及と志向:ヒトが{未来/過去}を想うとき

    久保圭, 仲村哲明, 臼田泰如, 荒牧英治, 荒牧英治

    言語処理学会年次大会発表論文集(Web)   20th  2014

    J-GLOBAL

  • 日本語カルテをアノテートする

    荒牧英治, 荒牧英治, 四方朱子, 島本裕美子, 久保圭, 宮部真衣, 大熊智子, 狩野芳伸, 狩野芳伸, 森田瑞樹

    言語処理学会年次大会発表論文集(Web)   20th  2014

    J-GLOBAL

  • マイクロブログから場所依存の様相記録を抽出する:“100ninmap”プロジェクトによる街歩きイベントの実施と応用

    宮部真衣, 北雄介, 久保圭, 荒牧英治

    言語処理学会年次大会発表論文集(Web)   20th  2014

    J-GLOBAL

  • 100-nin Map : Kyoto Map made by 100 people

    ARAMAKI Eiji, KITA Yusuke, NAKAMURA Tetsuaki, KUBO Kay, SHIMAMOTO Yumiko, USUDA Yasuyuki, MIYABE Mai

    IEICE technical report. Natural language understanding and models of communication   113 ( 338 ) 23 - 24  2013.11

     View Summary

    The "100-nin Map" is an event held on July 27, 2013. 54 people have participated in this event. They have posted about 2600 messages via a specially developed original application to create a newly generated map. This paper is to compare this original map with existing tourist maps (such as commercially available tourist maps and Twitter visualization based maps) to discuss the potential of collective intelligence.

    CiNii

▼display all

Works

  • 作文学習アプリ『いちまるとはじめよう!ことばむすび』

    京都大学, 日本漢字能力検定協会  Educational material 

    2022.01
    -
     

  • スマートフォン向け学習アプリ『漢検とニュース』

    京都大学, 日本漢字能力検定協会  Educational material 

    2019.04
    -
     

  • 日本語文章難易度判別システム (jReadability)

    李在鎬, 長谷部 陽一郎, 久保圭  Web Service 

    2016.04
    -
     

Awards

  • 2015年度 山下記念研究賞(共著者として受賞)

    2016.03   情報処理学会   音声認識による認知症・発達障害スクリーニングは可能か?:言語能力測定システム"言秤"の提案

  • ベストペーパー賞(共著者として受賞)

    2014.11   情報処理学会 グループウェアとネットワークサービスワークショップ2014   音声認識による認知症・発達障害スクリーニングは可能か?:言語能力測定システム"言秤"の提案

  • 学際研究着想コンテスト 優秀賞

    2014.08   京都大学   剥がれ落ちる言葉:特効薬のない認知症を言葉で治す斬新アプローチ

Research Projects

  • Construction and Analysis of Database for Emergency Nursing Education Based on Corpus Analysis

    Japan Society for the Promotion of Science  Grants-in-Aid for Scientific Research Grant-in-Aid for Scientific Research (C)

    Project Year :

    2015.04
    -
    2019.03
     

    HIRAO AKEMI, HARUNA Hiroko

     View Summary

    Students learn specialist knowledge in basic nursing education through lectures, exercises, and practical training. Textbooks and specialist books are used in lectures to understand the technical terms. However, in practical training, there is an addition of natural language that develops with the cultural background of a nursing facility for daily communications.
    In this study, we organized the technical terms related to emergency care and constructed a corpus database of emergency nursing care from nursing textbooks, specialized emergency nursing care books, as well as the nursing students’ practical training records that include natural language descriptions.

  • Development of a text readability assessment system to support reading activities in Japanese language education

    Japan Society for the Promotion of Science  Grants-in-Aid for Scientific Research Grant-in-Aid for Scientific Research (C)

    Project Year :

    2013.04
    -
    2016.03
     

    LEE Jaeho, HASEBE Yoichiro, KUBO Kei

     View Summary

    The research presented in this paper aims at advancing text readability studies for Japanese and devises a practical and useful system that contributes to Japanese language teaching, learning, and research. More specifically, utilizing leveled corpora mainly consisting of text from textbooks of Japanese, we produced the following formula to measure the readability level of a given text in six scales: X = {mean length of sentence * -0.056} + {proportion of kango * -0.126} + {proportion of wago * -0.042} + {proportion of number of verbs among all words * -0.145} + {proportion of number of auxiliary verbs * -0.044} + 11.724 (R2=.896) . The formula was tested against another set of leveled text in Japanese to prove its reliability. Lastly, the method was implemented in a computer system that calculates and produces the estimated level of a text via a web-based online interface.

Presentations

  • 100-nin Map : Kyoto Map made by 100 people

    ARAMAKI Eiji, KITA Yusuke, NAKAMURA Tetsuaki, KUBO Kay, SHIMAMOTO Yumiko, USUDA Yasuyuki, MIYABE Mai

    IEICE technical report. Natural language understanding and models of communication  The Institute of Electronics, Information and Communication Engineers

    Presentation date: 2013.11

    Event date:
    2013.11
     
     

     View Summary

    The "100-nin Map" is an event held on July 27, 2013. 54 people have participated in this event. They have posted about 2600 messages via a specially developed original application to create a newly generated map. This paper is to compare this original map with existing tourist maps (such as commercially available tourist maps and Twitter visualization based maps) to discuss the potential of collective intelligence.

 

Syllabus

▼display all

Teaching Experience

  • 日本語IV(聴解口頭)

    立命館大学  

    2019.04
    -
    Now
     

  • 日本語IV(総合)

    立命館大学  

    2019.04
    -
    Now
     

  • 日本語VI(キャリア日本語)

    立命館大学  

    2019.04
    -
    Now
     

  • 日本語VII(文章表現)

    立命館大学  

    2019.04
    -
    Now
     

 

Committee Memberships

  • 2019.06
    -
    Now

    日本語教育学会  広報委員

  • 2020.04
    -
    2022.03

    言語処理学会  大会プログラム委員

  • 2017.04
    -
    2018.03

    International Conference on Computer Assisted Systems for Teaching & Learning Japanese (CASTEL/J)  実行委員