Unveiling Emotions through Handwriting: A Data Analysis Approach

Detection of unfavourable feelings through routine tasks like handwriting is beneficial for encouraging well-being. This paper presents a study on emotion detection using handwriting data analysis. By examining a dataset consisting of 129 users and their performance on seven different tasks, we gained valuable insights into the relationship between emotional states and an individual’s handwriting. The study employed various techniques, including feature engineering, correlation analysis, and dimensionality reduction using Linear Discriminant Analysis and Principal Component Analysis. A Random Forest model is implemented to predict emotional states based on handwriting features. The results demonstrate promising accuracy in classifying emotional states. This study emphasizes the potential of handwriting analysis as a means to understand and detect emotions, laying the groundwork for further research in this domain.