EmotiDiet – Your Mood-Based Diet Buddy

• 

  Screenshot showing how EmotiDiet detects emotion from user input and recommends mood-based food.

Inspiration

Many people experience emotional eating, where their food choices are affected by stress, sadness, or mood swings. I wanted to build something small but meaningful using AI — a digital buddy that understands your feelings and gently suggests food to help improve your emotional state.

What It Does

EmotiDiet is an AI-powered app that predicts your emotional state based on free-text input and recommends food items that match or balance your mood. Whether you're feeling anxious, happy, sad, or stressed — EmotiDiet will guide you with personalized suggestions.

How I Built It

• Trained a machine learning model using the PIMA emotion dataset
• Built the web interface using Streamlit
• Deployed the app on Streamlit Cloud
• Managed code and versioning through GitHub

Challenges I Faced

• Training the model to understand emotional context in short, free-form text
• Designing a minimal, friendly interface that feels supportive
• Making sure the app works fast and is easy for anyone to use

What I Learned

• Improving ML model accuracy with text classification
• Real-time app deployment using Streamlit
• How a simple idea can have real-world usefulness and empathy

What's Next

• Add more emotion categories and smarter text understanding
• Suggest meal plans and not just food items
• Integrate voice input and possibly connect to fitness apps

Built With

• github
• joblib
• pandas
• python
• scikit-learn
• streamlit