FitMind: AI-Based Fitness and Mental Wellness Recommendation System with Chatbot Support

Abstract:

In today’s fast-paced lifestyle, maintaining both physical fitness and mental wellness has become a crucial but often neglected aspect of personal health. While there are numerous mobile and web applications addressing either fitness or mental health, very few offer a holistic, intelligent, and integrated solution. The FitMind system is a comprehensive web-based application designed to promote overall well-being through the power of artificial intelligence and machine learning.

The platform provides three major features: a personalized Fitness Planner, a Mental Health Tracker, and a Meditation & Wellness Advisor, all powered by robust ML models. The Fitness Planner Module predicts a user’s fitness category based on personal parameters like age, BMI, activity level, and lifestyle habits, and recommends a suitable workout plan and diet. The Mental Health Tracker Module leverages psychological screening scales like PHQ-9 (Patient Health Questionnaire-9), GAD-7 (Generalized Anxiety Disorder-7), and DASS-21 (Depression, Anxiety, and Stress Scale) to assess mental health conditions, and provides tailored advice for improving mental wellness. The Meditation & Wellness Module evaluates lifestyle parameters such as sleep quality, screen time, stress levels, and mindfulness score, then suggests guided meditations, breathing techniques, and daily wellness tips.

In addition, the system incorporates an AI-based Chatbot Module that uses natural language processing (NLP) and a Naive Bayes classifier to understand user queries and offer instant responses related to fitness, mental health, and meditation. Developed using Python, Flask, and MySQL, FitMind delivers an interactive and user-friendly experience with real-time recommendations and progress tracking.

Overall, FitMind serves as a smart virtual wellness assistant aimed at making mental and physical well-being accessible, personalized, and engaging for all users, especially students and working professionals. Its modular design also makes it scalable for future integration with wearable health devices and mobile platforms.

Introduction:

Mental and physical health are interconnected, and maintaining both is vital in today’s fast-paced lifestyle. The FitMind system is a comprehensive platform that analyzes fitness attributes, mental health symptoms, sleep patterns, stress levels, and mindfulness activities. It provides users with customized fitness plans, mental wellness evaluations, and guided meditation support. Additionally, the integrated AI chatbot acts as a virtual assistant to ensure accessibility, guidance, and user engagement through natural language interaction.

Problem Statement:

Most existing fitness and mental wellness applications work in silos, offering either workout suggestions or meditation separately. There’s a lack of integrated platforms that offer both, guided by AI and machine learning models. Users often find it challenging to navigate multiple apps for mental wellness, fitness, and consultation. Also, static rule-based systems lack personalization and adaptability.

Existing System and Disadvantages:

Existing apps provide:

• Basic fitness tracking
• Predefined meditation audios
• Text-based questionnaires without intelligence

Disadvantages:

• No integration between physical and mental health modules
• No predictive capability using ML models
• Lack of dynamic interaction through NLP
• One-size-fits-all approach

Proposed System and Advantages:

The proposed FitMind System includes:

• ML-based personalized fitness planner
• Mental health tracker using PHQ-9, GAD-7, DASS-21
• Meditation and wellness prediction
• AI chatbot powered by NLP for 24/7 guidance

Advantages:

• AI-driven recommendations
• Personalized based on real-time inputs
• Intelligent chatbot for instant support
• Visualization of predictions and improvement plans

Modules :

1. User Authentication Module
   • Handles secure user registration and login.
   • Validates inputs such as username, password, email, age, and gender.
   • Stores credentials securely using hashing techniques in MySQL.
   • Manages session handling to restrict unauthorized access.
2. Fitness Planner Module
   • Collects user inputs including age, BMI, activity level, and health goals.
   • Uses a Random Forest Classifier to predict an appropriate fitness plan (e.g., weight loss, muscle gain, maintenance).
   • Provides tailored workout routines and dietary recommendations.
   • Logs fitness results and plan feedback in the database for future tracking.
3. Mental Health Tracker Module
   • Enables users to complete PHQ-9 (Patient Health Questionnaire-9), GAD-7 (Generalized Anxiety Disorder-7), and DASS-21 (Depression Anxiety Stress Scale-21) assessments.
   • Applies machine learning classification to determine mental health status (e.g., normal, mild, moderate, severe).
   • Offers contextual guidance, such as therapy suggestions or self-care routines.
   • Visualizes mental health progress over time and saves results to user profiles.
4. Meditation & Wellness Module
   • Gathers data such as sleep hours, screen time, stress levels, and mindfulness habits.
   • Predicts wellness level (Good, Moderate, Poor) using a Random Forest model.
   • Based on prediction, dynamically provides appropriate guided meditation videos, breathing techniques, and daily wellness tips.
   • Encourages routine building and logs user wellness history.
5. AI Chatbot Support Module
   • Offers 24/7 conversational support using an NLP-based chatbot.
   • Classifies user queries using Naive Bayes trained on an intents JSON dataset.
   • Responds with helpful information across five domains: fitness, diet, mental health, meditation, and general assistance.
   • Improves user engagement, provides immediate feedback, and acts as a virtual wellness assistant.

Algorithms Used:

• Random Forest Classifier – For fitness, mental health, and wellness prediction
• Naive Bayes Classifier – For chatbot intent recognition
• CountVectorizer + LabelEncoder – For NLP-based classification

Software Requirements:

• Frontend: HTML5, CSS3
• Backend: Python, Flask
• Machine Learning: Scikit-learn, Numpy, Pandas, Joblib
• Database: MySQL
• Tools: VSCode / PyCharm, Postman, MySQL Workbench

Hardware Requirements:

• Processor: Intel i3 or higher
• RAM: 4 GB or more
• Storage: 2 GB minimum
• Webcam (optional for future upgrades)

Conclusion:

FitMind provides a unified AI-based solution to enhance user well-being through intelligent assessment and guidance. The use of ML models for fitness and mental health evaluation ensures personalized and accurate insights. The chatbot enhances user interaction and promotes continuous engagement and motivation for health goals.

Future Enhancement:

• Integrate real-time emotion recognition via webcam
• Add voice-based chatbot interaction
• Expand datasets with real user data for higher accuracy
• Integrate wearable device inputs (e.g., Fitbit, smartwatches)
• Enable multi-language chatbot interaction