【学术报告】Graph based Information Retrieval and More

发稿时间:2007-04-02浏览次数:2571

                                           LAMDA
                                   http://lamda.nju.edu.cn


题目:Graph based Information Retrieval and More

时间:4月6日(星期五),14:30-15:30

地点:蒙民伟楼404会议室

报告人:Dr. Dell Zhang,  
       Lecturer in Computer Science at Birkbeck College,
       University of London, UK

Abs:

Documents are often not independent but instead interrelated or interconnected with each other. More and more researchers in machine learning, data mining and information retrieval are now exploiting the explicit or implicit relationships among documents by modelling corpora as graphs. Some graph-theoretical methods originally proposed for the Web (e.g., PageRank and HITS) have been propagating to other types of document graphs, such as email networks and the blog osphere. Furthermore, there have recently been substantial advances in graph based manifold/semi-supervised learning and graph pattern mining. In this talk, I would like to give a brief overview about the usage of graph models, particularly spectral graph theory, for information retrieval, clustering, classification, and so on and so forth.


Bio:

Dell Zhang is a Lecturer in Computer Science at Birkbeck College, University of London. He got his Ph.D. from Southeast University in Jan 2002, and then worked as a Research Fellow in the Computer Science Programme of Singapore-MIT Alliance till he moved to the UK in Sep 2005. His research is on the theme of improving information retrieval/organization through machine learning or data mining. He has several publications in prestigious conferences such as KDD, WWW and SIGIR. He has served as a Demonstration Co-Chair for HLT/EMNLP-2005, and a PC member for SIGIR-2006, ECIR-2007, SIGIR-2007, etc.