Release Time:2019-03-11  Hits:

Indexed by:Journal Article

Date of Publication:2015-03-15

Journal:Journal of Computational Information Systems

Included Journals:Scopus、EI

Volume:11

Issue:6

Page Number:2199-2208

ISSN:15539105

Summary:Publication venues, which here refer to the journal or conference, are important and useful data in certain domains, such as publication retrieval, publication venue recommendation, authors name ambiguity and so on. However, for most cases, publication venues are used by conventional approaches based on text matching, which ignores the latent relations between publication venues. In this paper, we propose an approach for identifying the latent relations between publication venues based on latent semantic analysis (LSA) and we also visualize the relations by use of the results of our approach. The approach differs from others in that it uses LSA to mine the relations in latent semantic space rather than just uses simple text matching approaches and it also represents the relations in quantitative way. Firstly, the relations between authors and publication venues are represented by space vector model. Then non-negative matrix factorization (NMF) is used for constructing a latent semantic space via decomposition and dimensionality reduction. Finally the venue-venue relational model is calculated in the semantic space. The proposed approach is applied to DBLP dataset and the relations between publication venues are identified quantitatively. The latent relations between most venues in DBLP dataset are also visualized and the relations are represented, measured and searched in an intuitive way by our web system. Copyright ? 2015 Binary Information Press.