Construction and application of text classification model based on deep learning in online news

With the all-round development of digital information technology, online news has rapidly emerged as the main channel for the public to obtain information with its timeliness, universality and interactivity. However, in the face of the explosive growth of online news, how to find the information you are interested in or need quickly and accurately has become the key demand of current users. In this regard, in view of the current situation of online news retrieval and browsing functions, this paper will deeply explore the feasibility of applying the text classification model based on deep learning technology in the field of online news, and put forward a brand-new construction scheme of online news text classification model to improve the accuracy and expansibility of existing text classification methods. Practice has proved that the hybrid text classification model combining convolutional neural network (CNN) and long short-term memory (LSTM) has a better classification effect when classifying online news texts, with an average accuracy rate of 94.01%, which can provide more powerful support for subsequent news recommendation, public opinion monitoring and analysis applications.