基于EI-BA模型的文本情感分析

Text Sentiment Analysis Based on EI-BA Model

为深入挖掘微博文本的情感特征、充分考虑表情符号和标点符号对情感表达的作用,提出一种结合情感信息的深度神经网络模型。该模型利用余弦相似性计算表情符号和标点符号的情感得分,通过双向长短期记忆网络提取文本的上下文语义特征,引入注意力机制进一步提取与任务相关的情感特征,由Softmax分类器计算最终分类结果。模型应用在新型冠状病毒暴发初期的微博文本数据集上,结果表明其性能优于相关优异的基线方法。

In this paper,we propose a deep neural network model that combines emotion information to explore the emotional fea-tures of microblog text and fully consider the role of emoji and punctuation in emotion expression.The model uses cosine similarity to calculate the emotional scores of emoticons and punctuation,and extracts contextual semantic features of text through BiLSTN.Attention mechanism is introduced to further extract task-related emotional features,and the final classification results are calculated by Softmax classifier.The model is applied to the microblog text data set at the beginning of the new coronavirus outbreak,and the results show that the performance of the model is better than that of the baseline method.