Research on the Barrage Classification Algorithm for the Protection of Teenager on Video Websites

In order to avoid the adverse impact of barrages on teenagers on video website, the author realizes the layering, classification and grading of users' barrages by using the text classification algorithm of deep learning, guides the users' barrages and creates a good ecological environment of network language. The algorithm of barrage classification display needs to be updated in real time. In terms of data, we need to expand the sample of machine learning training in real time, in addition, the classification of language score should be further refined and accurate. In terms of technology, we can combine the latest development of psychology, communication and statistics to make the classification of barrages more accurate.