基于文本特征和语言知识的神经网络情感分类

Neural network sentiment classification based on text features and linguistic knowledge

为进一步提高文本情感倾向性分类效果,提出基于文本特征和语言知识融合的卷积神经网络模型MI-CNN。使用Word2Vec表示词语信息,将词性和情感词语等语言知识嵌入词向量中,将文本特征和语言知识融合到情感倾向性分类模型,经过参数优化提升文本情感倾向性分类模型的准确率。在数据集上进行实验,结果表明所提出的模型准确率达到93.0%,比文献中的基准模型取得了更好的分类效果。

In order to improve the performance of sentiment classification model,a convolutional neural network model MI-CNN based on text feature and language knowledge fusion is proposed in this paper.Firstly,Word2Vec is used to represent word information,and linguistic knowledge such as lexical and emotional words is embedded into word vector.Then text features and sentiment knowledge are fused into sentiment classification model.The accuracy is improved through parameter optimization.The result of experiments show that the proposed model achieves better performance than the benchmark model in literature,and the accuracy reach 93.0%.