基于情感词向量和BLSTM的评论文本情感倾向分析

情感倾向分析主要用于判断文本的情感极性，在商品评论、舆情监控等领域有着重要的商业和社会价值。传统的机器学习方法主要是浅层的学习算法，并不能很好地抽取文本中高层情感信息。针对该问题，提出了一种以组合了语义信息和情感信息的情感词向量作为输入的改进双向长短期记忆模型，通过构建语义和情感双输入矩阵，并在隐藏层加入情感特征抽取模块，来增强模型的情感特征表达能力。在数据集上的实验结果表明，与标准的BLSTM模型和传统机器学习模型相比，该模型能够有效提升文本情感倾向分析的效果。

Sentiment orientation analysis of review text based on sentiment word embedding and BLSTM

Sentiment orientation analysis is mainly used to judge the emotional polarity of the text, and has important commercial and social value in the fields of commodity review and public opinion monitoring. The traditional machine learning method is mainly shallow learning algorithm, and cannot extract the high-level emotional information in the text. Motivated by this, this paper proposed a BLSTM model with sentiment word embedding as input. Through building semantic and sentiment dual input matrix, and adding the emotion feature extraction module in the hidden layer, to enhance the emotion expression ability of the model. The experimental results show that compared with the standard BLSTM model and the traditional machine learning model, the model can effectively improve the effect of sentiment orientation analysis.