计算机应用研究2017，Vol.34Issue(9)：2696-2699,4.DOI:10.3969/j.issn.1001-3695.2017.09.029
一种改进EM算法的跨领域情感分类方法
Improved EM-based cross-domain sentiment classification method
摘要
Abstract
Supervised learning algorithm is currently the main method for text sentiment classification,and this method requires that the distribution of the training data set and test set is the same.But in practice,the distribution is always not in the same which will make the performance of the classification accuracy rate worse.In order to solve the above problems,this paper proposed a cross domain sentiment analysis method based on improved EM algorithm.Firstly,it generated a sentiment tendency reference list from multiple sources.And then used the improved EM algorithm to adjust the classification results of the target domain iteratively till the results could match the reference list.Finally,it took the experiments in the open dataset comparative with mainstream classification algorithms such as Na（i）ve Bayes and SVM.Experimental results show that the prooosed method can improve the accuracy of cross domain sentiment analysis to a certain extent.关键词
跨领域情感分类/EM算法/特征迁移Key words
cross-domain sentiment classification/EM algorithm/feature adaption分类
信息技术与安全科学引用本文复制引用
黄瑞阳,康世泽..一种改进EM算法的跨领域情感分类方法[J].计算机应用研究,2017,34(9):2696-2699,4.基金项目
国家科技支撑计划资助项目(2014BAH30B01) （2014BAH30B01）
国家自然科学基金创新群体资助项目(61521003) （61521003）
国家自然科学基金资助项目(61379151) （61379151）
