报告题目：Semantic-Driven Multi-Label Text Representation Learning

报告人：景丽萍 教授 北京交通大学 

报告时间：2021-01-11  09:00-10:00

腾讯会议ID：491 914 838

Abstract: Multi-label data, esp. multi-label textual data is ubiquitous in We-Media era. Comparing with the traditional single-label text, multi-label data contains complex syntactic and semantic structure. It is important and challenging to extract the discriminative information and mine the correlation among labels for improving the label prediction performance. In this talk, some recent work about multi-label text representation learning, label structure representation learning will be given.