Inspiration

The internet can be a wonderful place to learn and connect—but unfortunately, many kids face bullying online. We wanted to create a project that makes online spaces safer and kinder for everyone, especially young students.

What It Does

NiceBot is an AI-powered chatbot that gently detects negative, harmful, or inappropriate messages using sentiment analysis and keyword detection. Instead of punishing users, it encourages positive communication with empathetic replies like:

"⚠️ Try being nicer! We're all friends here."

It helps build a culture of kindness while subtly teaching kids how to communicate respectfully.

How We Built It

We used Python with Flask to build the backend and integrated TextBlob for basic sentiment analysis. The UI was made using responsive HTML and Tailwind CSS, optimized for both mobile and PC. The chatbot interface sends and receives messages in real-time using JavaScript’s fetch().

The logic adapts to rude words or negative sentiment and gives gentle feedback. The interface includes three pages:

  • Welcome screen with Start Chatting button
  • Chatbot page with interactive dialogue
  • Support form for help or feedback

Challenges We Faced

  • Making sentiment detection feel "human" instead of robotic.
  • Designing a UI that’s kid-friendly yet scalable across all devices.
  • Managing how AI responds to harsh or sarcastic language without being too strict or lenient.
  • Making everything simple enough for junior students to engage with.

What We Learned

  • How to use sentiment analysis and make AI emotionally aware.
  • Writing cleaner and responsive frontend code.
  • Building full-stack apps with frontend/backend integration.
  • How kindness can be designed into software.

What’s Next

  • Adding a database to track user conversations.
  • Creating a school-wide dashboard for reporting unsafe patterns.
  • Gamifying kindness points for young users.

Built With

Share this project:

Updates