A Deep Learning Approach for Predicting Personality Traits
Zainab F. Khan · 2023
One of the most effective science for determining a person's strengths and shortcomings is graphology, or handwriting analysis. Handwriting is a reflection of a person's personality. It reveals the writer's thoughts, emotions, and behaviors as well as the personality qualities one has and doesn't have. The personality can be profiled using Graphology in terms of traits like anger, self-esteem, pessimism, fear, honesty, work habits, social abilities, etc. Line spacing, margins, slant, letter size, text density, baseline inclination, and other handwriting elements are based on personality factors.The present paper proposes a deep convolution neural network (CNN) classifier for distinguishing eleven different lower case 't' alphabet patterns in graphology, each indicating different psychological attributes as willpower, self-esteem, motivation, and emotional stability. The suggested approach contributes to the area of graphology by providing a quick and accurate way to examine handwriting samples and identify the aforementioned characteristics. Eleven personality qualities are mapped onto all the characteristics. When evaluated on fresh, unseen handwritten documents, the suggested system had a categorization accuracy of at least 94%.