A network behavior evaluation method for improving interactive security in virtual network worlds

Abstract

ABSTRACTThe Internet is increasingly being used for socializing. In virtual network worlds (VNWs), security and trust evaluation mechanisms are weak. This study used a network behavior evaluation method to appraise the online behavior qualities and personal attributes of Internet users based on social computing. This paper first proposes a security interaction theory and factor model based on criminological theories. To understand importance of the factors affecting interactive security, we first analyze the background differences of 758 Internet users. We also design three interaction-security scales, which we use to evaluate the security values of other users. A network behavior evaluation method comprising three submethods is then developed. Finally, we present a fuzzy multiobjective decision inference (FMODI) method to integrate the three submethods to identify the optimal Internet user. There are 56 Internet users to participate in the method experiment, which include 19 valid samples regarding online entertainment, 22 samples regarding online shopping, and 15 samples regarding online social. We used quartiles to analyze these data, which we divided into four safety levels. This study provides a novel contribution to Internet safety evaluations. The security scores of the three network behavior evaluation submethods are used to warn or caution users of threats or identify users of potential threats, thereby preventing immediate damage.