Principal Research Manager, Microsoft Cognitive Services Research Group Ph.D., Computer Science Department, Stanford University M.S., Statistics Department, Stanford University B. Eng., Computer Science & Technology Department, Tsinghua University Email: A [at] B, where A=chezhu and B is microsoft.com [ Google Scholar | Books | Publications | Patents | Mentored Interns ]

Research Interests

 Knowledge Graph, Text Summarization, Task-Oriented Dialogue

 My recent work covers:

• How to increase factual correctness of abstractive summaries [ Paper ]

• Summarize multi-party meeting transcripts [ Paper ]

• Dialogue state tracking with slot-independent architecture [ Paper | Talk at Stanford HAI OVAL ]

• Conversational question answering [ Paper ]

Books

Machine Reading Comprehension: Algorithm and Practice 《机器阅读理解：算法与实践》 Chenguang Zhu China Machine Press (机械工业出版社) , 2020.03 [ Amazon.com | China-pub | jd.com | dangdang.com | tmall.com | Amazon.cn ] [ GitHub Code ]

Journal Publications

• Text Summarization
  文本摘要：浓缩的才是精华
  Chenguang Zhu, Nanshan Zeng
  Communications of the China Computer Federation (CCCF), 《中国计算机学会通讯》，Jul. 2020.

• Machine Reading Comprehension: How to Make Computer Understand Articles
  机器阅读理解：如何让计算机读懂文章
  Chenguang Zhu
  Communications of the China Computer Federation (CCCF), 《中国计算机学会通讯》，Feb. 2019.

Conference Publications

Knowledge Graph

• A Survey of Knowledge-Enhanced Text Generation
  Wenhao Yu, Chenguang Zhu, Zaitang Li, Zhiting Hu, Qingyun Wang, Heng Ji, Meng Jiang
  arXiv preprint arXiv: 2010.04389, 2020.

• JAKET: Joint Pre-training of Knowledge Graph and Language Understanding
  Donghan Yu*, Chenguang Zhu*, Yiming Yang, Michael Zeng
  (*: Equal contribution)
  arXiv preprint arXiv: 2010.00796, 2020.

Multimodal

• Semi-Supervised Speech-Language Joint Pre-Training for Spoken Language Understanding
  Yu-An Chung*, Chenguang Zhu*, Michael Zeng
  (*: Equal contribution)
  arXiv preprint arXiv: 2010.02295, 2020.

Summarization & Generation

• TED: A Pretrained Unsupervised Summarization Model with Theme Modeling and Denoising
  Ziyi Yang*, Chenguang Zhu*, Robert Gmyr, Michael Zeng, Xuedong Huang, Eric Darve
  (*: Equal contribution)
  Empirical Methods in Natural Language Processing (EMNLP), 2020.
  [ arXiv ]
  

• A Hierarchical Network for Abstractive Meeting Summarization with Cross-Domain Pretraining
  Chenguang Zhu*, Ruochen Xu*, Michael Zeng, Xuedong Huang
  (*: Equal contribution)
  Empirical Methods in Natural Language Processing (EMNLP), 2020.
  [ arXiv ]
  

• Injecting Entity Types into Entity-Guided News Generation
  Xiangyu Dong, Wenhao Yu, Chenguang Zhu and Meng Jiang
  arXiv preprint arXiv: 2009.13401, 2020.
  

• Boosting Factual Correctness of Abstractive Summarization with Knowledge Graph
  Chenguang Zhu, William Hinthorn, Ruochen Xu, Qingkai Zeng, Michael Zeng, Xuedong Huang, Meng Jiang
  arXiv preprint arXiv: 2003.08612, 2020.

• Mixed-Lingual Pre-training for Cross-lingual Summarization
  Ruochen Xu*, Chenguang Zhu*, Yu Shi, Michael Zeng, Xuedong Huang
  (*: Equal contribution)
  Asia-Pacific Chapter of the Association for Computational Linguistics (AACL), Suzhou, China, 2020.

• Make Lead Bias in Your Favor: A Simple and Effective Method for News Summarization
  [ Demo ]
  Chenguang Zhu, Ziyi Yang, Robert Gmyr, Michael Zeng, Xuedong Huang
  arXiv preprint arXiv: 1912.11602, 2019.

• Mind The Facts: Knowledge-Boosted Coherent Abstractive Text Summarization
  [ Poster ]
  Beliz Gunel, Chenguang Zhu, Michael Zeng, Xuedong Huang
  Conference on Neural Information Processing Systems (NeurIPS), Knowledge Representation & Reasoning Meets Machine Learning (KR2ML workshop), Vancouver, Canada, 2019.

Dialogue

• Few-shot Natural Language Generation for Task-Oriented Dialog
  Baolin Peng, Chenguang Zhu, Chunyuan Li, Xiujun Li, Jinchao Li, Michael Zeng, Jianfeng Gao
  Empirical Methods in Natural Language Processing (EMNLP), 2020.
  [ arXiv | Code & Demo ]
  

• Boosting Naturalness of Language in Task-oriented Dialogues via Adversarial Training
  Chenguang Zhu
  [ Talk ]
  Special Interest Group on Discourse and Dialogue (SIGdial), Boise, Idaho, 2020.

• Meta Dialogue Policy Learning
  Yumo Xu, Chenguang Zhu, Baolin Peng, Michael Zeng
  arXiv preprint arXiv: 2006.02588, 2020.

• Data Augmentation for Spoken Language Understanding via Pretrained Models
  Baolin Peng^∗, Chenguang Zhu^∗, Michael Zeng, Jianfeng Gao
  (*: Equal contribution)
  arXiv preprint arXiv: 2004.13952, 2020.

• Multi-task Learning for Natural Language Generation in Task-Oriented Dialogue
  [ Poster ]
  Chenguang Zhu, Michael Zeng, Xuedong Huang
  Empirical Methods in Natural Language Processing (EMNLP), Hong Kong, China, 2019.

• SIM: A Slot-Independent Neural Model for Dialogue State Tracking
  [ Talk at Stanford HAI OVAL | Poster ]
  Chenguang Zhu, Michael Zeng, Xuedong Huang
  Special Interest Group on Discourse and Dialogue (SIGdial), Stockholm, Sweden, 2019.

Textual Embeddings

• Filtered Inner Product Projection for Multilingual Embedding Alignment
  Vin Sachidananda, Ziyi Yang, Chenguang Zhu
  arXiv preprint arXiv: 2006.03652, 2020

• Parameter-free Sentence Embedding via Orthogonal Basis
  [ Code | Slides | Talk ]
  Ziyi Yang, Chenguang Zhu, Weizhu Chen
  Empirical Methods in Natural Language Processing (EMNLP), Hong Kong, China, 2019.
  

• Embedding Imputation with Grounded Language Information
  [ Poster ]
  Ziyi Yang, Chenguang Zhu, Vin Sachidananda, Eric Darve
  Association for Computational Linguistics (ACL), Florence, Italy, 2019.

Machine Reading Comprehension

• Learning to Attend On Essential Terms: An Enhanced Retriever-Reader Model for Open-domain Question Answering
  [ Code | Poster ]
  Jianmo Ni, Chenguang Zhu, Weizhu Chen, Julian McAuley.
  North American Chapter of the Association for Computational Linguistics (NAACL), Minneapolis, USA, 2019.

• SDNet: Contextualized Attention-based Deep Network for Conversational Question Answering
  [ Code ]
  Chenguang Zhu, Michael Zeng, Xuedong Huang.
  arXiv preprint arXiv: 1812.03593, 2018.

• FusionNet: Fusing via Fully-Aware Attention with Application to Machine Comprehension
  [ Code | Poster ]
  Hsin-Yuan Huang, Chenguang Zhu, Yelong Shen, Weizhu Chen.
  Sixth International Conference on Learning Representations (ICLR), Vancouver, Canada, 2018.

Transportation

• Reducing Inefficiencies in Taxi Systems
  Chenguang Zhu, Balaji Prabhakar.
  56th IEEE Conference on Decision and Control (CDC), Melbourne, Australia, 2017.

• Measuring the Pulse of a City Via Taxi Operation: Case Study
  Chenguang Zhu, Balaji Prabhakar.
  Transportation Research Board 96th Annual Meeting, Washington, D.C., 2017.

• Reducing Road Congestion Through Incentives: A Case Study
  Chenguang Zhu, Jia Shuo Yue, Chinmoy V. Mandayam, Deepak Merugu, Hossein Karkeh Abadi, Balaji Prabhakar.
  Transportation Research Board 94th Annual Meeting, Washington, D.C., 2015.
  Featured on The New York Times, The Wall Street Journal, International Business Times, Ars Technica and Stanford News

Social Networks

• Polling One's Friends: A Graph Theoretic View
  Chenguang Zhu, Hossein Karkeh Abadi, Balaji Prabhakar.
  53rd Annual Allerton Conference on Communication, Control, and Computing (Allerton), 2015.

• Information Diffusion and External Influence in Networks
  Seth A. Myers, Chenguang Zhu, Jure Leskovec.
  ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), 2012

Web Mining

• A Novel Click Model and Its Applications to Online Advertising
  Zeyuan Allen Zhu, Weizhu Chen, Tom Minka, Chenguang Zhu, Zheng Chen.
  ACM International Conference on Web Search and Data Mining (WSDM), 2010

• A General Magnitude-Preserving Boosting Algorithm for Search Ranking
  Chenguang Zhu, Weizhu Chen, Zeyuan Allen Zhu, Gang Wang, Dong Wang, Zheng Chen.
  ACM Conference on Information and Knowledge Management (CIKM), 2009

Optimization

• Inverse Time Dependency in Convex Regularized Learning
  Zeyuan Allen Zhu, Weizhu Chen, Chenguang Zhu, Gang Wang, Haixun Wang, Zheng Chen.
  IEEE International Conference of Data Mining (ICDM), 2009 Best Student Paper Award Runner-Up

• P-packSVM: Parallel Primal Gradient Descent Kernel SVM
  Zeyuan Allen Zhu, Weizhu Chen, Gang Wang, Chenguang Zhu, Zheng Chen.
  IEEE International Conference of Data Mining (ICDM), 2009

Patents

• Using machine comprehension to answer a question (US 20190156220)
  Chenguang Zhu, Hsin-Yuan Huang, Pengcheng He, Weizhu Chen, Yelong Shen, Zheng Chen.

• Conversational Virtual Assistant (US 20180232376)
  Chenguang Zhu, Weizhu Chen, Jianwen Zhang, Xuedong Huang, Zheng Chen.

• Caching Content Addressable Data Chunks for Storage Virtualization (US 20140280664)
  Sudipta Sengupta, Chenguang Zhu, Chun Ho Cheung, Jin Li, Abhishek Gupta.

Mentored Interns

I am very fortunate to have mentored and worked with talented interns.

• Donghan Yu (2020 summer), Carnegie Mellon University.

• Yu-An Chung (2020 summer), MIT.

• Yumo Xu (2020 spring), University of Edinburgh.

• Ziyi Yang (2018 summer and 2019 summer), Stanford University. Our paper on word embedding was published in ACL 2019. Our paper on sentence embeddings was published in EMNLP 2019. Our paper on unsupervised text summarization was published in EMNLP 2020.

• Beliz Gunel (2019 summer), Stanford University. Our paper on knowledge-boosted text summarization was published in NeurIPS 2019 workshop of KR2ML.

• Jianmo Ni (2018 summer), UC San Diego. Our paper on machine reading comprehension was published in NAACL 2019.

• Hsin-Yuan Huang (2017 summer), Caltech. Our paper on machine reading comprehension was published on ICLR 2018.

Services

• Program committee member of EMNLP 2019, ACL 2019, NeurIPS 2019.

• Program committee member of ACL 2018 Machine Reading and Question Answering workshop (MRQA).

Talks

• Knowledge Graph and Its Applications in NLP. Seminar at Department of Computer Science and Engineering, University of Notre Dame, 2020.09 [Link]

• Machine Reading Comprehension

  • Future of Computing Forum, Tsinghua University, 2020.05 [Video]

  • AI Express, Microsoft Online, 2020.05 [Video]

• Research Progress in Task-Oriented Dialogue. First Open Virtual Assistant Workshop, Stanford University, 2019.10 [Video]

• SDNet: Contextualized Attention-based Deep Network for Conversational Question Answering. Stanford NLP Seminar, Stanford University, 2019.01

• FusionNet: Fusing with Fully Aware Attention in Machine Reading Comprehension

  • Stanford Platform Lab Seminar, Stanford University, 2018.02

  • Guest lecture at EE392K, Stanford University, 2018.02

• Reducing Inefficiencies in Taxi Systems. Presented at IEEE CDC 2017, Melbourne, Australia, 2017.12

• Analysis and Modeling of Large-scale Transportation Systems

  • Guest lecture at EE392K, Stanford University, 2016.02

  • Presented at Google Research, Mountain View, CA, USA, 2016.02

  • Presented at Microsoft Research, Redmond, WA, USA, 2016.02

• Polling One’s Friends: A Graph Theoretic View. Presented at Allerton 2015, Monticello, Illinois, USA, 2015.09

• Reducing Road Congestion Through Incentives: A Case Study. Presented at TRB 2015, Washington DC, USA, 2015.01

• A General Magnitude-Preserving Boosting Algorithm for Search Ranking. Presented at ACM CIKM 2009, Hong Kong, China, 2009.11

Awards

• ACM International Collegiate Programming Contest (ICPC), World Finals 2012: 13th place (Representing Stanford University)

• UPE First Solution Award (For Solving the First Problem in ACM/ICPC World Final 2012) Photo at Award Ceremony

• Winner of Stanford Local Programming Contest, 2010, 2011

• Best Student Paper Award Runner-Up at IEEE International Conference of Data Mining (ICDM), 2009

• Yao Award (named after the Turing Award Laureate Andrew Yao), Tsinghua University

• Scholarship for excellent academic performance, Tsinghua University (awarded to top 2% students)

Miscellaneous

• National champion in US National Table Tennis Championships U2000 Division D, Dec. 2015

• Second Runner-up in men's singles table tennis match of Tsinghua University