40 likes | 132 Views
Summary Report, Fall 2010. Research: Exploiting Text-Rich Information Network for Topic Modeling and Object Clustering (Preparing for KDD’11) Cross-document Entity Resolution for News Articles Extract Information Network from News Articles Joint work:
E N D
Summary Report, Fall 2010 • Research: • Exploiting Text-Rich Information Network for Topic Modeling and Object Clustering (Preparing for KDD’11) • Cross-document Entity Resolution for News Articles • Extract Information Network from News Articles • Joint work: • One paper at ICDM’10 (with Heli, Jianbin, Jiawei, Peixiang) • One paper at CIKM’10 (with Jianbin, Heli, Jiawei, Yizhou) • Services: • PC : WSDM’11, NIPS-MLSC workshop • External Reviewer: ICDE’11, WWW’11, ACM TIST • Others • Courses: CS598CXZ Advanced Topics in IR, CS591 DM Seminar • Attend INARC APP kickoff meeting at CUNY
Exploiting Text-Rich Information Network for Topic Modeling and Object Clustering • Objective: To find a set of topics and cluster different types of objects simultaneously • Intuition: The topics of a document tend to be consistent with the topics of its authors and conference. Basic idea: Incorporate heterogeneous information network into topic modeling
Experimental Results With Prior Information