面向BSP-CNN的短文本情感倾向性分类研究

Research on Classification of Short Text Emotional Tendency for BSP-CNN

针对消费短文本评论中的情感倾向性分类问题,提出了一种BSP-CNN混合神经网络模型。模型先使用双向简单循环单元(BiSRU)对数据进行特征表示,再使用逐点卷积神经网络(P-CNN)进一步学习语义特征,并输出情感倾向性分类结果。实验结果表明,与传统的长短期记忆神经网络(LSTM)和卷积神经网络(CNN)相比,BSPCNN混合神经网络模型有效简化了计算,缩短了运行时间,并且在不同大小和不同文本长度的数据集上均能取得更高的F1值。

In view of the classification of emotional tendency in the short text comments on consumption, a BSP-CNN hybrid neural network model is proposed. The model first uses the Bidirectional Simple Recurrent Unit(BiSRU)to characterize the data, then uses Point-by-point Convolutional Neural Network(P-CNN)to further learn semantic features and output the results of emotional tendency classification. Experimental results show that compared with traditional Long Short-Term Memory neural networks(LSTM)and Convolutional Neural Networks(CNN), the BSP-CNN hybrid neural network model effectively simplifies calculation, shortens the running time, and obtains higher F1 socre on data sets of different sizes and text lengths.