Research on Garbage Barrage Recognition Method Integrating Self Attention Mechanism
The rise of online live streaming has brought various types of junk barrage.The traditional garbage barrage recognition method has a low recognition rate and cannot meet the current needs of garbage barrage recognition on live streaming platforms.This article analyzes the characteristics of garbage barrage text,introduces self attention mechanism,and combines ERNIE and TextCNN models to design an algorithm to improve the recognition rate of garbage barrage.Through comparative experiments,the rationality and feasi-bility of this method have been proven,providing a certain reference for future research in related fields.
classification of short textlive barrageTextCNN modelERNIE modelself attention mechanism
万方数据
维普
