An Empirical Model of Gender Based Human Trait Identification from Blogs

摘要

The growth of blogosphere has prompted researchers for knowledge discovery from the blogs. This paper presents a brief review on research works in this area and proposes an empirical model of gender and trait identification. The feature vector comprises of three elements viz., formal measure, gender preferential measure and stylistic measure. The various significant parts-of-speech found in the blogs are used in computing the features. The model is implemented in two stages viz. author gender identification and their personality trait identification. The model identifies gender based traits viz., Extrovert, Introvert and Thinker. The model is trained satisfactorily with textual data collected from blogs using adaptive network fuzzy inference system algorithm.