International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 04 | April 2024
p-ISSN: 2395-0072
www.irjet.net
Personality Prediction Using Handwriting Analysis Dr. H K Chethan1, Mohammed Ali Y M2, Suman Raj3, Syed Zeeshan4 , Shreyas Rajesh Pawar5 1Professor, Dept. of Computer Science and Engineering, Maharaja Institute of Technology Thandavapura,
Karnataka, India
2,3,4,5Students, Dept. of Computer Science and Engineering, Maharaja Institute of Technology Thandavapura,
Karnataka, India ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract – Understanding human personality traits plays a
Knowing personality features has been highlighted as a crucial instrument for matching individuals to occupations, assisting employers in the recruiting process, and assisting coaches in customizing instruction to fit various topic areas. Within the structure of psychological analysis, the knowledge acquired from the text's analysis is considered significant for designing a customized therapy strategy that engages the patient's mind.
pivotal role in various fields, including human resources, psychology and social sciences. In this project, we suggest a unique method for predicting personality leveraging deep learning algorithms applied to handwriting analysis. Handwriting being an uncommon attribute provides valuable insights into an individual's personality characteristics. The dataset comprises handwritten samples categorized into five personality traits: Anxious, Cooperative, Enthusiastic, Responsible and Openness. Leveraging Convolutional Neural Networks (CNNs), Artificial Neural Networks (ANNs), and Residual Neural Networks (ResNets), Our goal is to detect complicated patterns in handwriting photographs to predict personality traits accurately. Our methodology involves preprocessing handwritten images, including normalization and feature extraction, subsequently the development and training for models based on deep learning. We employ CNNs to capture spatial dependencies in handwriting patterns, ANNs for feature extraction and dimensionality reduction, and ResNets for handling deep network architectures effectively. A significant amount of testing and analysis is done to determine how well the suggested models work. We use criteria like recall, accuracy, precision, and F1-score to assess how well the models predict each personality feature Furthermore, we investigate ensemble methods for integrating forecasts from many models for improved precision and resilience. The results demonstrate promising performance in personality prediction, with the proposed approach achieving competitive accuracy rates across all personality traits. Furthermore, the models exhibit resilience to variations in handwriting styles and dataset sizes, highlighting their generalization capabilities.
In addition, detailed information collected in writing regarding good behaviour during a crime investigation can help solve the crime problem. In broad terms, the project aims to bring together historical practice and contemporary technology to offer a useful method for examining the complexity of human behaviour via the writing lens. Along with the growth of education, real-world applications that can improve every aspect of human life and humanity as whole are also desired results.
1.1 OVERVIEW To forecast personality characteristics based on handwriting characteristics, the project "Personality Prediction Through Handwriting Analysis" involves collecting a dataset of handwritten samples labelled with personality traits, extracting relevant characteristics from the handwriting, and using machine learning techniques.
1.2 PROBLEM STATEMENT While behavioural observations and subjective selfreport questionnaires are the mainstays of traditional personality evaluation methods, they are labour-intensive, prone to bias, and often inaccurate. Furthermore, these techniques frequently fail to convey the complexities and complex aspects of a person's personality. In contrast, handwriting analysis has a distinctive chance to obtain insights into personality traits; nonetheless, manual analysis necessitates skill and is labour-intensive. The goal of the problem statement is to create an automated system that can more accurately identify personality traits from handwriting than current approaches can.
Key Words: Handwriting Analysis, Personality Traits, CNN, ANN, ResNets
1.INTRODUCTION The importance of learning to write is to understand people and their behaviour in depth. This project looks at the distinctive characters through a review of texts in an effort to better understand human behaviour. Rather than relying on direct questions or observations, the research explored the ability of writing to reveal unconscious aspects of person’s behaviour. This project explores the historical and cultural origins of handwriting and aims to bridge traditional and modern techniques.
© 2024, IRJET
|
Impact Factor value: 8.226
1.3 OBJECTIVE By first compiling a varied and labelled dataset of handwritten samples with associated personality qualities,
|
ISO 9001:2008 Certified Journal
|
Page 1536