AI-POWERED COGNITIVE PROFILING AND HIDDEN TRAIT DISCOVERY SYSTEM FOR PERSONALITY MAPPING.

Abstract

Personality assessment has historically relied on psychology-based questionnaires (MBTI) such as this one which is widely used in education, recruitment and counseling. Psychometric testing today is rapidly being adopted in India, especially in HR, EdTech, and mental wellness sectors such as so many large organizations use personality tests and these are in use as the basis for the development of the present study. In the research, we are designing an automated system for predicting MBTI personality traits from user-generated text to be based on artificial intelligence to provide scalable, real-time and unbiased personality profiles without lengthy questionnaires or manual evaluation. In existing systems XGBoost is used with text embeddings and linguistic features to classify personality traits. This method is effective for tracking the interaction of features and nonlinear relationships but is not sufficient for sequential dependencies and contextual nuances in natural language. To overcome this problem, we have proposed a Bidirectional Long Short-Term Memory (BiLSTM) model that learns contextually and sequentially from text. The system uses Sentence Transformers to generate high-dimensional semantic embeddings that capture context, meaning, and writing style. Then it feeds these embedded data into the BiLSTM network and they process the text in both forward and backward directions to understand linguistic dependencies. This deep learning method helps the model detect subtle personality signals in unstructured text. The BiLSTM-based architecture addresses the challenges of questionnaire and machine learning approaches and reduces self-report bias, enhances context understanding, and provides real-time personality prediction at scale. It is a robust and efficient tool for automated personality profiling and is suitable for real-world work with recruitment, psychological counseling, and personalized learning systems.