Distinction of AI Comments Based on HMM

Abstract

“AI water army†has implicitly coming, its development will beset people's life. In order to distinguish AI comments and human comments, this paper constructs an distinction framework: it is suggested that the sample should be taken from a large amount of network comments by using Reservoir sampling and then use HMM to build a mathematical model to predict the state of the comment-AI or human. Due to the limitations of the data, this paper adopts a simplified HMM algorithm to distinguish the comment state. The validity of the framework is verified by the results obtained. If more data is available, the accuracy of the model will be greatly improved.