基于Renyi熵的文本情感分析  

Text Sentiment Analysis Based on Renyi Entropy

作　　者：张冠东 姜荣 ZHANG Guandong;JIANG Rong(School of Mathematics,Physics and Statistics,Shanghai Polytechnic University,Shanghai 201209,China)

机构地区：[1]上海第二工业大学,数理与统计学院,上海201209

出　　处：《微型电脑应用》2023年第12期16-18,共3页Microcomputer Applications

基　　金：教育部人文社会科学研究青年基金项目(22YJC910005)。

摘　　要：为了能够通过文本语句的关键词理解整个语句的褒贬含义,提出一种改进型Renyi熵,通过对所抽取的文本关键词可能涵盖的不同词性的褒贬程度进行分析,有效地对该语句的褒贬进行分类。该方法使得文本褒贬识别方式接近于人类思维方式,与其他几个现有的模型相比,其褒贬的分类准确度和精确度均有所提高。同时,该方法也为文本情绪分析领域的研究提供了新思路。In order to understand the positive or negative meaning of the whole statement through the keywords of the text statement,an improved Renyi entropy is proposed to classify texts by analyzing the positive or negative meaning of several possible characteristics of a certain word.This method makes the recognition of positive or negative text close to the way of human thinking.Compared to several other existing models,the accuracy and precision of the improved Renyi entropy have been improved in positive or negative text classification.Meanwhile,this method also provides a new idea for the research of text sentiment analysis.

关 键 词：RENYI熵 文本分类 情感分析 词类型 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]