A Computer Identity Authentication Technique Based on User Behavior

Identity authentication based on physical characteristics has high security, which needs the support of specific external devices. In the past two years, the message that the authentication based on physical characteristics has been decoded has been repeatedly transmitted. No matter whether the security factor is high or low, password authentication and physical characteristics belong to the category of one-time authentication. After the authentication is passed, all permissions are granted to the account. In this paper, the method of big data and machine learning is adopted to design and implement the user authentication model based on behavior characteristics, which belongs to the authentication method that the user’s behavior content in the operation process is taken as the analysis object. In addition, the algorithm is optimized and innovated to assign different weights to the features with different stability and cluster the data, which greatly improves the efficiency and rationality of the authentication algorithm. As an implicit authentication, behavioral authentication features security, low cost, and sustainability can improve account security.