Chinese Text Sentiment Analysis Based on Improved Convolutional Neural Networks

Aiming at the shortcomings of the existing depth learning methods in text sentiment classification task, a Chinese text emotion classification algorithm with three-channel convolution neural network is proposed. The textual Pinyin is proposed as the corpus of feature extraction for the first time. The PCCNN (Phoneticize Channel Convolutional Neural Network)model is trained by extracting Pinyin char features and traditional word and char features, which improves the model's ability to extract features and improves the classification performance of the model. Experiments show that the performance of the algorithm is not only easier to implement than the traditional algorithm, but also the classification performance is more accurate. At the same time, the sentiment of each category is strictly superior to the single-channel convolutional neural network algorithm.