Predicting Conscientiousness through Semantic Analysis of Facebook Posts

Proceedings of the International AAAI Conference on Web and Social Media, 7

Volume

Vol. 7 No. 2 (2013): ICWSM Workshop Technical Report WS-13-01 (Computational Personality Recognition — Shared Task)

Computational Personality Recognition

Here we present a method for detecting an individual's level of conscientiousness based on an analysis of the content of their Facebook status updates. Our model is based on the identification of semantic evidence of facets related to conscientiousness; an individual's belief of their control over events around them and their goal orientation. The model achieves a correlation of r=.27 on a subset of the Facebook data published for the myPersonality workshop, with an accuracy of 58.13% for detecting if an individual is above or below the median and 68.03% for those outside of one standard deviation. While we take a narrow approach and identify only one personality trait, the general methodology of directly looking for evidence of traits in an individual's utterances is applicable to discovering models for all of the personality traits.

10.1609/icwsm.v7i2.14471