文章基本信息

标题：Efficient Text Classification Model Based on Improved Hyper-sphere Support Vector Machine with MapReduce and Hadoop
本地全文：下载
作者：Manisha Bhonde ; Prof. Pramod Patil
期刊名称：International Journal of Advanced Research In Computer Science and Software Engineering
印刷版ISSN：2277-6451
电子版ISSN：2277-128X
出版年度：2013
卷号：3
期号：8
出版社：S.S. Mishra
摘要：the concept of text classification is process of sorting out the text documents automatically in predefined classes. There are many algorithms are presented for the automatic text categorization. Those algorithms are representing the bags or words and hence processing the large number of features. For the feature extraction semantic analysis is commonly used, removing the text representation errors caused by synonyms and polysemes, and hence removing the vector dimensions. Recently we have studied the Hyper-sphere-SVM (HS-SVM) as the recent machine learning method for text classification, this method later suppressed by Improved HS-SVM (IHS-SVM) by gaining more accuracy and efficiency. In this paper we are extending and investigating the IHS-SVM method by addition the parallel processing methods of text classification such as MapReduce and Hadoop. In case of IHS-SVM, additional to existing method of text classification new decision-making method based on concentration is presented for enhancing the classification of texts in overlapping regions. The practical evaluation of IHS-SVM with and without MapReduce as well as Hadoop is presented in this paper
关键词：Text Classification; machine learning methods; SVM; HS-SVM; IHS-SVM; MapReduce; Hadoop