Automated Personality Profiling from Handwriting using Machine Learning
Aritra Bhattacharyya, Mansi Wadhwa, Mohammed Adil Shaikh · 2024
Handwriting analysis, also known as graphology, has been traditionally used as a method for assessing personality traits. Various characteristics of handwriting, like letter size, slant, and pressure applied, are believed to provide insights into an individual’s personality. Graphologists typically follow a systematic approach, analyzing handwriting samples by examining various aspects of the writing style and comparing them to established interpretations or psychological theories. This paper presents a novel machine learning architecture developed to infer personality traits from handwritten text samples. Five different handwriting features have been extracted, and feature extraction has been performed using an Ensemble (Random Forest) technique. The proposed model achieved an accuracy of 87.6% in predicting personality traits according to the Big Five personality dimensions. These results indicate the usefulness of the approach in automating handwriting analysis for personality assessment.