Multi-Label Classification of English Social Media Texts based on Deep Learning Algorithm

Multi-label classification of English social media texts has become an important task in natural language processing (NLP). Traditional classification methods have shortcomings in dealing with the semantic complexity of texts and the dependencies between labels. This paper proposes an efficient multi-label classification method for English social media texts based on the CNN+LSTM algorithm. First, the social media texts are preprocessed, and then a hybrid model based on a convolutional neural network (CNN) and a long short-term memory (LSTM) network is used to capture the local features and long-distance dependencies in the text. The method introduces a multi-label classification loss function to optimize the performance of the model in multi-label scenarios. During the model training process, a large-scale social media dataset is used and compared with the traditional MLP. The experimental results show that the proposed method has significant advantages in multi-label classification tasks, with an average accuracy of 85.3%. The recall rate and F1 value of the model are also improved by an average of 13.2% and 0.15, respectively, indicating that the method is superior in dealing with scenarios with unbalanced labels and close relationships between labels, and can provide strong support for the accurate classification and information retrieval of social media texts.