A Machine Learning Approach for Predicting Personality Traits

Handwriting analysis, also known as graphology, is one of the most effective ways to get insight into a person's personality as well as their strengths and shortcomings. It reveals the thoughts, feelings, and actions, as well as the characteristics of the writer's personality and those that are lacking. Graphology allows for the profiling of the personality in terms of a wide variety of characteristics, including but not limited to anger, self-esteem, optimism, fear, openness, productivity, social skills, etc. The numerous characteristics of a person's handwriting, such as line spacing, inclination, slant, text density, letter size, margins, baseline, and so on, are used to determine personality traits. The proposed system includes prepossessing, feature extraction, and, finally, training of a machine learning classifier called a Support Vector Machine (SVM). This classifier is hyperparameter tuned to learn major components and intricate details of handwriting samples, such as the size, slant, pressure, baseline angle, and word spacing of the handwriting samples.Following the completion of the normalization step, the extracted features are subsequently labeled according to the thresholds that were set through experimentation. In the end, all of the characteristics are mapped into a total of eight different personality traits. When tested against handwritten materials that were not seen beforehand, the proposed method achieved a classification accuracy of 96%.