Researh And Implementation Of Socail Bots Detection System Based On Feature Fusion

In recent years,with the progress of Internet technology and artificial intelligence technology,more and more people browse and express their views on social platforms.Therefore,social robots,a program based on artificial intelligence technology,gradually appear on social networks.Many social robots create false information,fabricate online reviews,spread false news about political candidates,and maliciously attempt to control and manipulate public emotions.As more and more intelligent social robots can fully imitate real users,the detection technology of social robots has important research significance.This paper studies the behavior of social robots and the detection technology of social robots.Based on the detection technology of social robots based on artificial intelligence,the main work is as follows:1.This paper proposes the method and main structure of feature combination,and divides the detection of social robot based on feature combination into feature extraction layer,feature combination layer and output layer.2.Based on the structure of feature combination for social robot detection,this paper carries out pairwise feature combination in three dimensions of time,theme and emotion,and designs the specific implementation method of each layer structure in each combination.3.This paper also uses micro service technology to design social robot detection system,which is suitable for the method structure of feature Fusion.The innovations of this paper include:1.The structure of social robot detection using feature Fusion is proposed.2.LDA and mogrifier LSTM are introduced to detect social robots based on the combination of temporal and thematic features.3.Introducing Vader,ARIMA time series algorithm and mogrifier LSTM to detect social robots with the combination of time and emotion characteristics.4.Design a scoring system to detect the combination of theme and emotional features of social robots.