Yanhua Chen, Wayne State University, USA
Lijun Wang, Wayne State University, USA
Ming Dong, Wayne State University, USA

Links

Session:
Springer Link:

Abstract

In order to derive high quality information from text, the field of text mining has advanced swiftly from simple document clustering to co-clustering with words and categories. However, document co-clustering without any prior knowledge or background information is a challenging problem. In this paper, we propose a Semi-Supervised Non-negative Matrix Factorization (SS-NMF) framework for document co-clustering. Our method computes new word-document and document-category matrices by incorporating user provided constraints through simultaneous distance metric learning and modality selection. Using an iterative algorithm, we perform tri-factorization of the new matrices to infer the document, category and word clusters. Theoretically, we show the convergence and correctness of SS-NMF co-clustering and the advantages of SS-NMF co-clustering over existing approaches. Through extensive experiments conducted on publicly available data sets, we demonstrate the superior performance of SS-NMF for document co-clustering.