Open AccessA Personality Prediction Method of WEIBO Users based on Personality Lexicon
Author(s) -
Yuanyuan Feng,
Kejian Liu
Publication year - 2021
Publication title -
natural language processing
Language(s) - English
Resource type - Conference proceedings
DOI - 10.5121/csit.2021.112312
Subject(s) - lexicon , personality , construct (python library) , word embedding , computer science , perspective (graphical) , big five personality traits , artificial intelligence , natural language processing , psychology , embedding , social psychology , programming language
Personality is the dominant factor affecting human behavior. With the rise of social network platforms, increasing attention has been paid to predict personality traits by analyzing users' behavior information, and pay little attention to the text contents, making it insufficient to explain personality from the perspective of texts. Therefore, in this paper, we propose a personality prediction method based on personality lexicon. Firstly, we extract keywords from texts, and use word embedding techniques to construct a Chinese personality lexicon. Based on the lexicon, we analyze the correlation between personality traits and different semantic categories of words, and extract the semantic features of the texts posted by Weibo users to construct personality prediction models using classification algorithm. The final experiments shows that compared with SC-LIWC, the personality lexicon constructed in this paper can achieve a better performance.