Shreyash Rai

About the project

Glutamate

BERT fine-tuned for sentiment, served live inside a Chrome extension.

  • Python
  • BERT
  • Transformers
  • Chrome extension
REPOSITORY

Overview#

Glutamate is a simple browser extension designed to analyze user sentiments in comments and provide feedback. Built as part of a hackathon project, this extension leverages AI/ML models to detect the emotional tone of user comments, helping content creators optimize engagement and decision-making.

This was built in a 24-hour hackathon “Code of the Phoenix” - E-Summit 2025 at IIIT Naya Raipur

Problem Statement#

Human language expresses a wide range of emotions, and detecting sentiment in online comments can be challenging due to:

  • The complexity of tone, sarcasm, and irony.
  • The need for real-time emotion recognition to provide instant feedback.

Objectives#

  • Detect whether a comment is positive, negative, or neutral.
  • Provide appropriate feedback based on sentiment to enhance user experience.
  • Ensure ease of use through a simple browser extension.

Features#

  • Emotion Detection: Identifies emotions such as happiness, sadness, and anger.
  • Real-Time Feedback: Provides personalized feedback to users based on detected sentiment.
  • Context Awareness: Understands the tone and context of comments.

How to Install Extension#

  1. Go to Extension page on your browser.
  2. Make sure developer mode is turned on
  3. Click Load Unpacked from the sidebar
  4. Select the SentimentExtension folder
  5. Click Load and it will be installed on your browswer successfully.

How to Use#

  1. Install the extension.
  2. Highlight a comment on any website.
  3. Right-click and select ‘Analyze Sentiment’.
  4. View the sentiment analysis result and suggested feedback.
  5. Alternatively, manually enter text in the provided textbox and press the analyze button.

Technical Details#

  • Dataset: IMDB comment dataset (used for training the AI model).
  • AI Model: BERT (Bidirectional Encoder Representations from Transformers) for sentiment classification.
  • Backend: Flask (Python) for AI model integration.
  • Frontend: JavaScript for Chrome extension UI.

Future Improvements:#

  • Expanding the dataset to include platforms like Twitter and YouTube.
  • Enhancing sarcasm and complex tone detection.
  • Multilingual sentiment analysis.
  • Multi-modal recognition (text, images, and voice analysis).
  • Scalable AI solutions for real-time enterprise feedback.

Contributors#

  • Shreyash Rai (Team Lead)
  • Shourya Vaidhya Jain(Developer)

Conclusion#

Glutamate enhances sentiment analysis for online comments, helping users gain insights into emotional tone and improve engagement. Future iterations will focus on more advanced AI techniques and a broader dataset to improve accuracy and context understanding.

Thank you for checking out Glutamate! 🚀