计算机软件新技术国家重点实验室 

摘 要：

       Automatically generating questions from text has many application in diverse fields. Many existing work focuses on generating only questions without concerning itself with identifying answers. Moreover, our analysis shows that handling rare words and generating the most appropriate question given a candidate answer are challenges still facing existing approaches. In this talk, I will present our ongoing work on generating question-answer pairs from text using deep neural networks. I will present a number of techniques we have been developing. Using standard evaluation metrics as well as human evaluation, our results show significant improvements in the quality of questions generated by our techniques over the state-of-the-art. If time permits, I will also discuss our work on visual question answering (VQA).

报告人简介：

      Yuan-Fang Li is a Senior Lecturer at Faculty of Information Technology, Monash University. Yuan-Fang received his Bachelor of Computing (Honours) and PhD from School of Computing, National University of Singapore (NUS). His research interests include knowledge graphs, knowledge representation and reasoning, ontology languages, natural language processing, and software engineering.

时间：2月28日  10:00-11:00