Our group’s research interests broadly lie in Data Mining and Artificial Intelligence, with emphasis on text mining and understanding, various kinds of question answering, network analysis, and human behavior understanding.
We aim to develop advanced computational frameworks that enable users to effectively query big data, collaborate with each other, and acquire knowledge for decision making in various domains such as healthcare, education, and business.
- 11/2018: Our recent work on “Interactive Semantic Parsing for If-Then Recipes via Hierarchical Reinforcement Learning,”was accepted to AAAI’19 (acceptance rate: 16.2%)! Please check out our interactive semantic parser here!
- 11/2018: Our work on “Answer Identification from Product Reviews for User Questions by Multi-task Attentive Networks” was accepted to AAAI’19 (acceptance rate: 16.2%)!
- 07/2018: Our paper “A Comprehensive Study of StaQC for Deep Code Summarization,” an extension study of our <natural language question, code snippet> dataset mined from Stack Overflow in WWW 2018, was accepted in KDD 2018 Deep Learning Day (SPOTLIGHT)!
- 12/2017: Work by Ziyu Yao on mining large-scale high-quality <natural language question, code snippet> pairs got accepted in WWW 2018. This is our very first work along the research line to facilitate Computer Science for All. Please check our datasets, code, and paper here!
- 06/2017: Work by Jie Zhao on answer triggering got accepted in EMNLP 2017. Code is now available here. Please let Jie or me know if you have any suggestions!
- 01/2017: Got tired of simple question answering? Try out our characteristic-rich question set! Any suggestions are welcome.
- 12/2016: Our paper “Reliable Medical Diagnosis from Crowdsourcing: Discover Trustworthy Answers from Non-Experts” was accepted in WSDM 2017
- 09/2016: Our paper “An Augmented LSTM Framework to Construct Medical Self-diagnosis Android” was accepted in ICDM 2016
- 07/2016: Our paper “On Generating Characteristic-rich Question Sets for QA Evaluation” was accepted in EMNLP 2016
Webpage construction credit to Ruoqi Liu