Personality prediction is an important application area of affective computing. The process of assessing the personality of users is proposed in this project using questionnaires answered by the user and classifying the personality according to MBTI traits. This project proposed a framework in which some questions are to be answered by the user. Every question has an option in the form of agree or disagree and then the model predicts the personality of the user from the 16 MBTI traits with a description of personality trait based on the questions answered. This project used deep learning approach. Recurrent Neural Network(RNN) along with Long Short-Term Memory units (LSTM) is used to develop this project. Accuracy of the proposed model achieved is 64.5 which could overcome the average accuracy of personality prediction classifier made while previously it was 55% using RNN. This model which makes use of Questionaries’ can be used as a feedback system in academics and organisation for improving the quality of life of students and employees.

Deep learning model that predicts a person’s dominant personality trait (Extraversion, Agreeableness, Conscientiousness, Neuroticism, or Openness) using sleep pattern data such as sleep duration, distance, app usage, and average sleeping time.

This project predicts Big Five personality traits (Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism) using machine learning and deep learning. The notebook covers data preprocessing, training multiple models (Logistic Regression, Random Forest, XGBoost, LSTM), evaluating results, and explaining predictions with SHAP and LIME.