基于自联想记忆与卷积神经网络的跨语言情感分类  

Cross-lingual Sentiment Classification Based on Auto-associative Memory and Convolutional Neural Networks

作　　者：刘娇[1] 崔荣一[1] 赵亚慧[1] LIU Jiao;CUI Rongyi;ZHAO Yahui(Intelligent Information Processing Lab.,Department of Computer Science and Technology, Yanbian University,Yanji,Jilin 133000,China)

机构地区：[1]延边大学计算机科学与技术学院智能信息处理研究室,吉林延吉133000

出　　处：《中文信息学报》2018年第12期118-124,共7页Journal of Chinese Information Processing

基　　金：国家语委"十三五"科研规划项目(YB135-76);延边大学外国语言文学世界一流学科建设科研项目(18YLPY13)

摘　　要：该文提出了一种以商品评论为对象的基于语义融合的跨语言情感分类算法。该算法首先从短文本语义表示的角度出发,基于开源工具Word2Vec预先生成词嵌入向量来获得不同语言下的信息表示;其次,根据不同语种之间的词向量的统计关联性提出使用自联想记忆关系来融合提取跨语言文档语义;然后利用卷积神经网络的局部感知性和权值共享理论,融合自联想记忆模型下的复杂语义表达,从而获得不同长度的短语融合特征。深度神经网络将能够学习到任意语种语义的高层特征致密组合,并且输出分类预测。为了验证算法的有效性,将该模型与最新几种模型方法的实验结果进行了对比。实验结果表明,此模型适用于跨语言情感语料正负面情感分类,实验效果明显优于现有的其他算法。A cross-linguistic sentiment classification algorithm based on semantic fusion is proposed for product reviews.First,information of different languages is generated by the open-source tool Word2 Vec in advance.Then,the auto-associative memory relationship is proposed to extract the cross-lingual document semantic,according to statistical relevance of word vector between different languages.Local perception and weight sharing techniques of convolutional neural networks are applied to amalgamate of complex semantic expression in auto-associative memory model,so as to generate the phrase features of different lengths.The dense combination of high-level semantic features is learned by deep neural network for all languages,which paves the way for classification predictions.It is demonstrated that,for positive and negative sentiment classification of cross-lingual sentiment corpus,the proposed model is much more effective than other existing algorithms.

关 键 词：跨语言情感分类 自联想记忆 词共现 卷积神经网络 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]