基于Adaboost算法的主客观句分类

Classification of Subjective and Objective Clauses Based on Adaboost Algorithm

中文语句的语义表达复杂,使用单一分类器进行主客观句分类的效果一般。该文提出一种基于Adaboost算法进行主客观句分类的方法。首先介绍了主客观分类的研究现状及一般流程;然后引入Adaboost集成学习算法,并针对算法的退化现象进行了相关的改进;最后在实验中使用了词汇线索特征和2-POS特征作为输入对短文本进行分类,结果表明Adaboost在主客观分类应用中效果良好。

Because the semantic expression of Chinese sentences is complex,the effect is not good to classify subjective and objective sentences by using a single classifier. In this paper,a method based on Adaboost algorithm is proposed for the classification of subjective and objective sentences. The research status and general process of the subjective and objective classification are introduced. Then Adaboost algorithm is introduced and some related improvements are done to avoid the degeneration phenomenon. Finally,lexical feature and 2-POS feature are taken as an input to classify short texts. The experimental results show that Adaboost has a good effect in the application of the subjective and objective classification.