您好,欢迎访问东北财经大学管理科学与工程学院!(请使用IE8及以上浏览器访问,360浏览器请开启极速模式) English| 中文| 加入收藏

IS Research Seminar:桑盛田

IS Research Seminar:桑盛田
IS Research Seminar   一种基于图表示学习与深度学习的隐含知识发现方法
 
 
报告人:桑盛田
时  间:2019年4月17日8:20-9:20
地  点:劝学楼 425
 
【报告人简介】
     桑盛田博士毕业于大连理工大学计算机学院计算机软件与理论专业,博士就读期间曾在美国斯坦福大学计算机学院访问学习。研究领域包括数据挖掘、文本挖掘、机器学习与自然语言处理等。其相关研究成果发表在BMC Bioinformatics,IEEE International Conference on Bioinformatics and Biomedicine, BioMed research international,BMC Medical Informatics and Decision Making等国内外重要期刊和会议。
 
【Abstract】
Literature-based discovery is a kind of data mining method, which focuses on generating new knowledge by combining what is already known in literature. In this study, we proposed a new literature-based discovery method which incorporates the knowledge graph embedding and deep learning method for mining implicit knowledge from literature. Firstly, the method builds a knowledge graph by exploiting the relations extracted from literature. Then, the graph data are converted into a low dimensional space by leveraging the knowledge graph embedding methods. After that, a recurrent neural network model is trained by the known entity associations which are represented by graph embedding. Finally, the learned model could be used to discover hidden knowledge for the entities of interest. The experimental results show that our method could not only effectively discover hidden knowledge by mining literature, but also could provide the corresponding explanations for the discovered associations. In addition, this method could be a supplementary method in multiple research areas for discovering novel knowledge from unstructured data.