A Mobile Security-Related Behavior Prevention Model Based on Speech Personality Traits

摘要

With the continuous development of mobile Internet, the damage due to mobile security incidents is motivating system and security application to adopt protective mechanisms. Recent research has begun to focus on the correlation between personality traits and security-related behavior such as email phishing response. In practical instances of this security domain knowledge, acquiring the user's personality accurately and conveniently in a mobile environment is desired. In this paper, considering the high efficiency of speech personality assessment, a prevention model using speech personality traits is proposed to guarantee customized security services. Our model is based on multilayer perceptron and prosodic features selection to acquire user speech personality traits. Furthermore, by combining obtained personality traits and security-related behavior domain knowledge, a prevention strategy is proposed. The proposed prevention strategy, which design is based on the correlation between user personality profile and security behavior weaknesses, is used to provide customized security prevention services. The results showed that the proposed model can achieve superior performance than other state-of-the-art speech personality assessment method and guarantee customized security-related behavior prevention support according to personality traits. The assessment time on mobile terminals is reduced to 19.2% of the traditional questionnaire method.