Text Classification Retrieval Based on Complex Network and ICA Algorithm

Journal of Multimedia

关键词: Collocation;    Independent Component Analysis;    ICA Clustering Algorithm;    Text Classification Retrieval;    Complex Networks;   

Others  :  1017367
DOI  :  10.4304/jmm.8.4.372-378

【 摘 要 】

With the development of computer science and information technology, the library is developing toward information and network. The library digital process converts the book into digital information. The high-quality preservation and management are achieved by computer technology as well as text classification techniques. It realizes knowledge appreciation. This paper introduces complex network theory in the text classification process and put forwards the ICA semantic clustering algorithm. It realizes the independent component analysis of complex network text classification. Through the ICA clustering algorithm of independent component, it realizes character words clustering extraction of text classification. The visualization of text retrieval is improved. Finally, we make a comparative analysis of collocation algorithm and ICA clustering algorithm through text classification and keyword search experiment. The paper gives the clustering degree of algorithm and accuracy figure. Through simulation analysis, we find that ICA clustering algorithm increases by 1.2% comparing with text classification clustering degree. Accuracy can be improved by 11.1% at most. It improves the efficiency and accuracy of text classification retrieval. It also provides a theoretical reference for text retrieval classification of eBook

【 授权许可】

   
@ 2006-2014 by ACADEMY PUBLISHER – All rights reserved.