Personality analysis through handwriting recognition

Abstract

The uniqueness of an individual can be identified according to one’s behavior, qualities, skills, characteristics, social connectivity, knowledge which all resemble to a single term called, “Personality.” There are ‘n’ number of ways to identify personality of an individual like judging the face gesture, way of speaking or communication, dressing style, questionnaire format, public behavior, etc. To identify what kind of a personality an individual has, traits can be classified based on individual’s handwriting. And this can be identified using the field ‘Graphology’. Graphological tests are the most effective method to measure (a candidate’s) ‘fit’ or match, for a position being applied for. According to research, it is proven that person’s brain is only 15-20% active while communicating whereas the same brain is active around 85-90% when writing. Hence, analysis of handwriting stands at par to describe and analyze the personality traits of individuals. The paper showcases a landscape of handwriting recognition strategies deployed for personality analysis, some of which involve the use of traditional neural networks like Convolution Neural Network and Recurrent Neural Network. The results presented in this paper obtained by implementing the neural networks such as a Long Short-Term Memory or a Fuzzy Logic along with possible classifier sets are significant enough in terms of higher accuracy and recognition.

Subject Classification:

• Primary 68T10
• Secondary 62H30

Keywords:

• Convolutional Neural Network
• Long Short-Term Memory
• Fuzzy Logic
• Multi-Modeling
• Deep Learning
• Optical Character Recognition