1Mrs.L. SHIRISHA, 2MANISHAKUMARI, 3V. AISHWANTH, 4T. THARUN 03 05 Mental health disorders such as depression anxietyand stress have become significant global concernsdue to rapid lifestyle changes increased socialpressure and technological dependency Earlydetection of such conditions remains challengingbecause many individuals hesitate to seekprofessional help or remain unaware of their mentalhealth status This project presents a machinelearningbased system MindPredictor designed toclassify mental health conditions using supervisedlearning algorithms by analyzing usergeneratedtextual data particularly from social mediaplatforms The system leverages natural languageprocessing techniques including textpreprocessing tokenization stemminglemmatization and sentiment analysis to extractmeaningful insights from user input Sentimentpolarity and subjectivity scores are computed andused as features for classification The processeddata is then fed into machine learning models suchas Nave Bayes and hybrid classifiers to distinguishbetween depressive and nondepressive states Thesystem evaluates model performance using metricslike accuracy precision recall and confusionmatrix analysis to ensure reliable predictionsResults demonstrate that machine learningalgorithms can effectively identify patterns relatedto mental health conditions and provide accurateclassification outcomes The proposed systemoffers a costeffective scalable and accessiblesolution that can assist in early mental healthassessment It does not replace professionaldiagnosis but acts as a supportive tool forawareness and preliminary screening Byintegrating artificial intelligence with healthcarethe system contributes to improved mental healthmonitoring timely intervention and reduced socialstigma associated with psychological disorders Copyright (c) Journal of Science Engineering Technology and Management Science. All rights reserved 2026