基于语义嵌入深度学习的Web文本情感分类  

Construction of Complex Network Text Sentiment Analysis Model Based on Deep Learning

作　　者：彭云[1] 万红新[2] Peng Yun;Wan Hongxin(School of Computer and Information Engineering,Jiangxi Normal University,Nanchang 330022,China;School of Mathematics&Computer Science,Jiangxi Science&Technology Normal University,Nanchang 330038,China)

机构地区：[1]江西师范大学,江西南昌330022 [2]江西科技师范大学,江西南昌330038

出　　处：《黑河学刊》2021年第5期122-128,共7页Heihe Journal

基　　金：江西省高校人文社科项目“融合语义关联和文本降维的社交媒体大数据关键主题提取研究”(项目编号:JC19121);江西省教育厅科技项目“基于深度学习的文本多粒度情感语义提取研究”(项目编号:GJJ201127)。

摘　　要：互联网用户数量的迅猛增涨,为网络提供了大量的文本数据,对这些海量文本进行有效的情感分类,是自然语言处理领域需要解决一个难题。Web文本的情感反映了网络用户利用文字抒发的针对某对象或事物不同的情感表示,情感极性的分类可以是褒奖的、中立的或批评的。由于中文Web文本的语义表达结构和形式复杂多样,一般模型难以实现基于语义理解基础上的情感分类,提出了基于语义嵌入深度学习的情感分类模型。将语义提取、语义嵌入和深度学习相结合,实现Web文本在复杂语境下的有效情感分类。The rapid increase in the number of Internet users has provided the Internet with a large amount of text data.Effective sentiment classification of these massive texts is a problem that needs to be solved in the field of natural language processing.The sentiment of Web text reflects the different emotional expressions of a certain object expressed by web users through text,and the classification of sentiment polarity can be commendatory,neutral or critical.Due to the complex and diverse semantic expression structure and forms of Chinese Web text,it is difficult for general models to realize sentiment classification based on semantic understanding.A sentiment classification model based on semantic embedding deep learning is proposed.Combining semantic extraction,semantic embedding and deep learning,it realizes effective sentiment classification of Web text in complex context.

关 键 词：WEB文本 语义嵌入 情感分类 深度学习 

分 类 号：C37[社会学]