Distinguishing Personality Recognition and Quantification of Emotional Features Based on Users' Information Behavior in Social Media

Abstract

The purpose of this paper is to explore the emotional composition, psychological characteristics, and the consistency between information behavior and attitude of social media users, and to provide reference for online public opinion monitoring, topic detection, and emotional situation evaluation. Based on big-five personality theory and self-difference theory, this paper takes 12,151 Twitter texts during Hurricane Maria as the analysis objects, extracts the personality characteristics of the texts based on convolution neural network, and analyzes the subjectivity and emotional polarity of the texts by Python. Based on the experimental results, this paper analyzes the psychological characteristics and information needs reflected by social media users' information behavior in disaster environment and further verifies and expounds the reasons for the inconsistent information behavior and attitude of social media users in disaster environments.