Hits:

Indexed by:期刊论文

Date of Publication:2013-08-15

Journal:Journal of Computational Information Systems

Included Journals:EI、Scopus

Volume:9

Issue:16

Page Number:6343-6350

ISSN No.:15539105

Abstract:This paper proposes a bilingual biomedical topic space model in which both Chinese and English abstract are represented by improved Latent Dirichlet Allocation with SVD and NMF decomposition smoothing methods. The improved LDA-based method is used to calculate the relationship between query and document and combine the results of different matrix factorization. A set of different dimension models is set up, with the help of which we can achieve bilingual cross-language indexing. The experimental results show that our method can improve the retrieval accuracies effectively. ? 2013 Binary Information Press.