基于权值算法的中文情感分析系统研究与实现

为了能够更加准确地对语句结构进行划分、对语句表达的内容进行判断,提出了一种全新的基于权值的计算算法,在完善中文分词的基础上对语句进行情感分析。首先利用中文分词算法对句式结构进行分割,然后依据词性对词库进行扩展,词库对句式中干扰词汇进行过滤,最后利用全新的权值计算算法对语句情感进行准确分析。经有效测试结果表明,情感分析准确率较高,并广泛适用于网络舆情分析等应用中。

Study and implementation of Chinese language emotion analysissystem based on weight algorithm

In order to separate the sentence structure and to judge the content of the statement more accurate, this paper put forward a new methodology of weight calculation based on the foundation of Chinese words segmentation and then analyzed the emotion of sentences. First, it adapted Chinese words segmentation algorithm to separate the sentence structure and then appended word dictionary according to the part of speech of words. Next, the methodology filtered disturbing words in sentences using word dictionary. Finally, the methodology took advantages of the new weight calculating algorithm to analyze the emotion of sentences accurately. Several validation tests approve that the result of the analysis has high accuracy and can be widely used in many applications such as public opinion analysis and so on.