About Me

Welcome to my homepage. I am broadly interested in artificial intelligence, with a particular focus on natural language processing, AI agents, voice agents, retrieval-augmented generation systems.

Education

• 2020.4 - 2023.3: Ph.D., Tokyo Institute of Technology

  • Supervisor: Prof. Manabu Okumura

• 2018.4 - 2020.3: M.E., Tokyo Institute of Technology

  • Supervisor: Prof. Hiroya Takamura

• 2012.9 - 2016.7: B.E., Beijing Language and Culture University

  • Supervisor: Prof. Dong Yu

Work Experience

• 2023.07 - current: Data Scientist, Asurion Japan, Tokyo
• 2022.04 - 2022.09: Applied Scitienst Intern, Amazon Alexa AI, Cambrdige, UK

  • Mentor and manager: Tom Ayoola, Andrea Pierleoni

• 2019.11 - 2020.02: Research Intern, Google Research, Mountain View, U.S.

  • Hosts: Amir Fayazi, Abhinav Rastogi

• 2017.02 - 2018.03: R & D Engineer, Sogou Inc., Beijing, China
• 2015.08 - 2017.02: R & D Intern/Engineer, Baidu Inc., Beijing, China

  • Intern: 5/2015-07/2016, Full-time: 7/2016-02/2017

Publications

• Yukun Feng, Amir Fayazi, Abhinav Rastogi and Manabu Okumura, ”Efficient Entity Embedding Con- struction from Type Knowledge for BERT”, Findings of The Asia-Pacific Chapter of the Association for Computational Linguistics (AACL-IJCNLP Findings 2020) [code]

• Yukun Feng, Chenlong Hu, Hidetaka Kamigaito, Hiroya Takamura and Manabu Okumura, Improving Character-Aware Neural Language Model by Warming Up Character Encoder under Skip-gram Architecture, The International Conference Recent Advances in Natural Language Processing (RANLP 2021) [code] [slide].

• Yijin Xiong, Yukun Feng, Hao Wu, Hidetaka Kamigaito, Hiroya Takamura and Manabu Okumura, Fusing Label Embedding into BERT: An Efficient Improvement for Text Classification, In Findings of the Association for Computational Linguistics (ACL-IJCNLP 2021 Findings)

• Chenlong Hu, Yukun Feng, Hidetaka Kamigaito, Hiroya Takamura and Manabu Okumura, One-class Text Classification with Multi-Modal Deep Support Vector Data Description, The European Chapter of the Association for Computational Linguistics (EACL 2021).

• Yukun Feng, Chenlong Hu, Hidetaka Kamigaito, Hiroya Takamura and Manabu Okumura, A Simple and Effective Usage of Word Clusters for CBOW Model, The Asia-Pacific Chapter of the Association for Computational Linguistics (AACL-IJCNLP 2020) [code] [slide].

• Yukun Feng, Hidetaka Kamigaito, Hiroya Takamura and Manabu Okumura, A Simple and Effective Method for Injecting Word-level Information into Character-aware Neural Language Models, The SIGNLL Conference on Computational Natural Language Learning 2019 (CoNLL 2019) [code] [poster].

• Yasufumi Taniguchi, Yukun Feng, Hiroya Takamura and Manabu Okumura, Generating Live Soccer-Match Commentary from Play Data, Thirty-Third AAAI Conference on Artificial Intelligence (AAAI 2019).

• Yukun Feng, Dong Yu, Jian Xu and Chunhua Liu, Semantic Frame Labeling with Target-based Neural Model , Sixth Joint Conference on Lexical and Computational Semantics (*SEM 2017, ACL workshop).

• Yukun Feng, Yipei Xu and Dong Yu, An end-to-end approach to learning semantic frames with feedforward neural network, NAACL 2016 Student Research Workshop.

• Yukun Feng, Qiao Deng, and Dong Yu, BLCUNLP: Corpus Pattern Analysis for Verbs Based on Dependency Chain, International Workshop on Semantic Evaluation (SemEval-2015, NAACL Workshop).