UNIVERSITY OF YAMANASHI > English > “DeepMet: A Reading Comprehension Paradigm for Token-level Metaphor Detection”, Chuangdong Su, Fumiyo Fukumoto, Xiaoxi Huang, Jiyu Li, Rongbo Wang, Zhiqun Chen Proc. of the Second Workshop on Figurative Language Processing, pp. 30-39, 2020./研究生院硕士课程工学专业双学位课程二年级学生苏传东 在计算语言国际会议比喻表现研讨会比喻检测任务竞赛中获得最优秀奖

“DeepMet: A Reading Comprehension Paradigm for Token-level Metaphor Detection”, Chuangdong Su, Fumiyo Fukumoto, Xiaoxi Huang, Jiyu Li, Rongbo Wang, Zhiqun Chen Proc. of the Second Workshop on Figurative Language Processing, pp. 30-39, 2020./研究生院硕士课程工学专业双学位课程二年级学生苏传东 在计算语言国际会议比喻表现研讨会比喻检测任务竞赛中获得最优秀奖

07/28/2020 English

  On July 9th, Mr. Su Chuandong, a student of master’s double degree program between Computer Science and Engineering Course, University of Yamanashi and School of Computer Science and Technology, Hangzhou Dianzi University, School of Computer Science and Technology won the best award in the metaphor detection shared task at the Second Workshop in Figurative Language Processing, Association for Computational Linguistics (ACL’20). ACL is the top conference in the natural language processing research field. The goal of the second shared task on metaphor detection is to identify, at a word level, all content-word metaphors in a given text. The shared tasks include two data sets of VU Amsterdam metaphor corpus and ETS Corpus of Non-Native Written English, and two tasks of verbal metaphor and all part-of-speech metaphor, with a total of metaphor detection that is complex, difficult and time-consuming. After nearly two months of hard work, our team achieved the best result in four metaphor detection tracks, beating the teams of Peking University, Cambridge University, Washington University, and other internationally renowned universities. Our related system description paper was published in ACL Second Workshop on Figurative Language Processing, and we could discuss it with experts in the field. Mr. Su Chuandong comments that “I am honored to have been selected as an excellent paper, and I am grateful to all of you for the guidance and cooperation that provided me with such joy. Thank you!”

The Second Workshop on Figurative Language Processing

　令和2年7月9日（星期四），举行了自然语言处理最难的国际会议“Annual Meeting of the Association For Computational Linguistics”，在比喻表现研讨会比喻检测任务竞赛中，研究生院硕士课程双学位课程2年级的苏传东（指导教师：福本文代教授、李吉屹助教、Xiaoxi Huang教授）开发的比喻检测软件DeepMet击败了北京大学、剑桥大学、华盛顿大学等14个团队，荣获最优秀奖。
　该竞赛共准备了4种自然语言处理任务，参加者利用附有答案的公开数据，在2个月内开发出比喻检测处理软件。之后，应用软件进行数据测试，与其他参赛团队就测试精度进行竞赛。
　苏同学开发的DeepMet，系统地整理了比喻检测所需要的各种语言知识，通过使用深度学习有效地进行学习，实现了高精度的检测。
　苏同学说：“非常荣幸得到最优秀奖。能够获得这样的奖项，多亏了一直以来指导我的老师和研究室的各位。我衷心地感谢您们。”

第2回比喻表现研讨会