Hits:

Indexed by:会议论文

Date of Publication:2016-07-05

Included Journals:EI、CPCI-S、Scopus

Page Number:193-198

Key Words:knowledge; naive bayes; Classification; mobile application (APP); Vector Space Model (VSM)

Abstract:In recent years, with the gradual development of mobile Internet technology, the number of mobile applications increases dramatically. Users facing numerous mobile applications are often caught off guard. It is necessary to automatically classify the applications according to the applications' information, so as to recommend appropriate applications to users. However, the text information directly obtained from APPs is vague and limited, to address this problem, this paper proposes to use Web knowledge to enrich the APPs' contextual information and takes a full advantage of Vector Space Model (VSM), Naive Bayes (NB) and Latent Dirichlet Allocation (LDA) to analyze t-he combined APPs' information in order to achieve a more accurate classification results. Experimental results show that the proposed method outperforms the traditional classification methods.