An Intelligent Classification Method for Online Resource Data of College Language Teaching Based on Deep Reinforcement Learning

Abstract The author starts from the classification of online teaching resources, and optimizes the current classification effect of online teaching resources through deep reinforcement learning algorithm. Taking language teaching in colleges and universities as an example, feature extraction is carried out through the characteristics of online educational resources using deep reinforcement learning. Subsequently, the deep reinforcement learning algorithm is optimized and the intelligent classification model of teaching resources based on DRML is constructed. Taking text data and image data in language online teaching resources as examples, the classification performance of DRML model is compared with other classical classification models to verify the classification performance of DRML model. The classification results of DRML are evaluated to determine the user’s satisfaction with it. The text classification performance of DRML classification model in this paper is better than other classification models. Classification using labeled semantic features outperforms the use of sentiment features and dynamic text classification outperforms static text features. The DRML model’s image classification accuracy is no less than 88%, which is a better performance than the best existing model. More than 75% of the users agreed/strongly agreed with the DRML model of this paper for classifying teaching resources, and the DRML model gained a high level of satisfaction.