Skype: billyuchenlin

Bill Yuchen Lin

"We were language’s magpies by nature, stealing what ever sounded bright and shiny."

I am a senior undergrad at Shanghai Jiao Tong University, majoring in Computer Science as a member of IEEE Honor Class. My advisor is Prof. Kenny Q. Zhu. In this past summer, I worked with Prof. Wei Lu as a visiting fellow of StatNLP Group in Singapore. Currently, I am taking my research internship at Microsoft Research Asia. ​My research interests include Natural Language Processing, Information Extraction, Knowledge Base, Social Networks, Computational Social Science, Data Mining, etc.

I am applying to CS PhD programs in Fall 2018. [resume]  


Shanghai Jiao Tong University, Shanghai, China

B.Sc. in Computer Science (Expected)
Sept. 2014 -- Jun. 2018
Advisor: Prof. Kenny Q. Zhu
Member of IEEE Honor Class

IEEE Honor Class is an elite program at SJTU directed by Prof. Xinbing Wang, with the aim of nurturing scientists in computer science, electrical and electronic technology, and information science based on MIT’s educational model.

Research Experience

Microsoft Research Asia, Beijing, China

Research Intern
Sept. 2017 -- Present
Focus: Semantic Parsing, Question Answering

StatNLP Group, Singapore

Visiting Fellow, supervised by Prof. Wei Lu
Jun. 2017 -- Sept. 2017
Focus: Domain Adaptation, Named Entity Recoginition, Knowledge Graph Representation

ADAPT Lab, Shanghai, China

Research Assistant, supervised by Prof. Kenny Q. Zhu
May. 2016 -- Present
Focus: Text Mining, Knowledge Discovery, Relation Extraction, Spatiotemporal Data Mining, etc.

IIoT Research Center, Shanghai, China

Research Assistant, supervised by Prof. Luoyi Fu and Prof. Xinbing Wang
Mar. 2017 -- Jun. 2017
Focus: Temporal Social Network Aanalysis

Research Projects

Under Review

[1] Topic: Transfer Learning for Named-Entity Recoginition
Supervisor: Prof. Wei Lu
[hide more]

Major Contribution: (1) we proposed a novel, lightweight transfer learning approach for cross-domain NER with neural networks, which does not require re-training on data from the source domain and can be potentially applied to other cross-domain structured prediction tasks; (2) through extensive empirical analysis as well as ablation study, we demonstrated the effectiveness of our approach, reporting better results over existing transfer methods. This work was submitted to NAACL-HLT 2018 (first author).

[2] Topic: Cross-Cultural Differences in Social Media.
Supervisor: Prof. Kenny Q. Zhu and Prof. Seung-won Hwang;
Collaborator: Frank F. Xu
[hide more]

In this paper, we present a lightweight yet effective framework for mining cross-cultural differences and similarities from social media, and evaluate it on two proposed novel tasks: (1) mining cross-cultural differences of named entities; (2) finding similar terms for slang terms across languages. The proposed framework can potentially become a building block for cross-cultural/lingual social media analysis. Extensive experiments show that our framework substantially outperforms a number of baseline methods on both tasks. This work was submitted to NAACL-HLT 2018 (first author). A preliminary work with little common content will be presented as a non-archivial paper at KBCOM@WSDM 2018, which instead utilizes image search results as resources.

[3] Topic: Commonsense Knowledge Graph Representation.
My Bachelor Thesis Proposal
Supervisor: Prof. Kenny Q. Zhu, Prof. Wei Lu
Collaborator: Yan Zhang

[4] ExtRA: A Framework for Extracting Review Aspects from Online Customer Feedback.
Zhiyi Luo, Bill Y. Lin, Frank F. Xu, Hanyuan Shi, Shi Feng, Kenny Q. Zhu
in revision for ICDE 2018   


* means equal contribution.

[5] Commonsense LocatedNear Relation Extraction.
Frank F. Xu*, Bill Y. Lin*, Kenny Q. Zhu
NIPS 2017 Workshop on Automated Knowledge Base Construction (AKBC@NIPS 2017)

[6] Multi-channel BiLSTM-CRF Model for Emerging Named Entity Recognition in Social Media.
Bill Y. Lin*, Frank F. Xu*, Zhiyi Luo, Kenny Q. Zhu
EMNLP 2017 Workshop on Noisy User-generated Text (W-NUT@EMNLP 2017)

[7] Dynamic Detection of Communities and Their Evolutions in Temporal Social Networks.
Yaowei Huang*, Jinghuan Shang*, Bill Y. Lin*, Luoyi Fu, Xinbing Wang
to appear in the Proceedings of AAAI 2018 (Student Poster)

[8] Transfer Learning for Traffic Speed Prediction: A Preliminary Study.
Bill Y. Lin*, Frank F. Xu*, Eve Q. Liao, Kenny Q. Zhu
to appear in the Proceedings of AAAI 2018 Workshop on Smart IoT, short paper

[9] Cross-region traffic prediction for China on OpenStreetMap.
Frank F. Xu, Bill Y. Lin, Qi Lu, Yifei Huang, Kenny Q. Zhu
in Proceedings of the 9th ACM SIGSPATIAL, International Workshop on Computational Transportation Science (IWCTS@SIGSPATIAL 2016)


0) call me Bill if you feel Yuchen is not a good name; call me Yuchen if you think Bill is not a good name. I agree with both of you!!
1) BiLL = Beauty in Language Learning
2) I'm a huge fan of American situation comedy series, such as Modern Family, How I Met Your Mother, Friends, The Big Bang Theory, etc. :)
3) My video essay for application to LTI@CMU.

Last Update: Dec. 24, 2017