报告题目：Dining with Neural Networks for Natural Language Processors

报告人：严睿 博士

时 间： 5月31日（周二）9：00-10：30

地 点： 北京大学计算机科学技术研究所1层106会议室

Abstract: In recent years, deep neural networks (DNNs, also known as deep learning) have made significant improvement in Natural Language Processing. DNNs are highly automated learning machines; they can extract underlying abstract features of data automatically by exploring multiple layers of non-linear transformation. In this talk, we invite every natural language processor to have a "dinner" with and take a closer look at deep neural networks. We briefly introduce the people and recent success in deep learning. Then we share the experience using DNNs for classification (ranking) tasks as well as generation tasks. For the ranking tasks, we illustrate with the deployment of DNNs in human-computer conversation systems, which is regarded as one of the most hardcore problems in artificial intelligence. For the generation tasks, we introduce how we apply DNNs in natural language generations with constraints. We hold an open discussion on where are we going with DNNs in the future.

Bio:Dr. Rui Yan is now a senior researcher in Baidu. He obtained the doctorate degree from Peking University. He is now also an adjunct professor and guest advisor in Central China Normal University and Central University of Finance and Economics. Rui Yan has a broad interest in real world problems related to natural languages, text information, social networks, web application, scientific literature, and multimedia. Rui's research focuses on Natural Language Processing, Information Retrieval, Machine Learning and Artificial Intelligence. More specifically, he is now conducting research into conversation systems, natural language generation, text mining and knowledge management. Till now, he has published more than 45 papers in various venues including ACL, SIGIR, IJCAI, WWW, KDD, and CIKM, etc.