Application of Emotion Classification of Digital Education Reviews Based on Convolutional Neural Network

In the information-based education environment, as a key link of teaching feedback, digital education review is becoming increasingly important. However, in the face of huge and unstructured text data, colleges and universities can only rely on a large number of manual reading, classification and statistical operations to process it, which is difficult to meet the practical application needs. In view of the current situation, this paper will deeply explore the application feasibility of artificial intelligence and machine learning technology in this field, and put forward a set of emotional classification model of digital educational comments based on convolutional neural network, aiming at improving the efficiency and accuracy of digital educational comments processing. Practice has proved that the whole model is built on the basis of CNN-Bi-LSTM, in which CNN is mainly responsible for extracting local features of the text and reducing the feature dimension, while Bi-LSTM focuses on capturing the time dependence and two-way context information of the sequence data. The combination of the two enables the model to use both local and global information, so as to output more accurate processing results in the task of emotion classification, and provide support for subsequent teaching management and decisionmaking.