Back to Catalog

Build a Negativity Filtering Web Extension with GenAI

BeginnerGuided Project

Enhance your web development and JavaScript skills by building a smart browser extension that utilizes the latest generative AI techniques. In this beginner friendly project you'll build a browser extension powered by LLMs and BERT to transform negative content. Through hands-on experience, you’ll learn how these language models process human text and discover their typical use cases. This project is the ideal launchpad for developers looking to integrate AI into their applications. Enroll now to expand your technical skills by building a tool that improves online experiences.

Language

  • English

Topic

  • Web Development

Skills You Will Learn

  • JavaScript, LLM, Web Development, NLP, Artificial Intelligence, BERT

Offered By

  • IBMSkillsNetwork

Estimated Effort

  • 30 minutes

Platform

  • SkillsNetwork

Last Update

  • June 30, 2025
About this Guided Project

Project Demo and Introduction

In this project, you will develop a browser extension that uses toxicity detection to analyze and filter text on a webpage. This tool will provide a simple solution to the widespread hate and negativity that permeates the Internet. The project will use a pre-trained BERT model (ToxicBERT) and an LLM (Meta's Llama 3) for text processing. The BERT model will detect toxicity past a certain threshold and the LLM will transform the text if that threshold is exceeded. Text processing will be integrated into a FastAPI app, making it easy to interface with via the JavaScript extension. Additionally, there will be steps for testing and deployment to use the extension on your local device. You will be guided through the full software development cycle, from design to build to launch. This comprehensive project will not only strengthen your programming skills but also provide hands-on experience with deploying AI in a practical application. By the end, you’ll have a working extension that leverages the power of natural language processing to address a common problem on the Internet.

A Look at the Project Ahead

By the end of this project you will:
  • Understand how BERT and LLMs process language in a nutshell
  • Use a pre-trained BERT model to detect negativity in web-scraped content
  • Prompt an LLM (Llama 3) to transform the text into a positive alternative
  • Build a FastAPI backend to wrap your text-processing pipeline
  • Build a simple JavaScript extension to scrape and modify the DOM
  • Test and debug your API within the CloudIDE
  • (Optional) Load and test the extension locally in your browser

What You'll Need

  • A solid foundation in JavaScript, Web APIs and Web Development
    • Familiarity with using the console to log and debug
  • A basic understanding of Python classes and Object Oriented Programming
  • A basic familiarity with APIs and how they are interfaced
  • A basic understanding of Git actions and commands
  • (Optional) Firefox or Chrome browser for testing

Instructors

Joshua Zhou

Data Scientist

I like building fun and practical things.

Read more

Contributors

Karan Goswami

Data Scientist

I am a dedicated Data Scientist and an AI enthusiast, currently working at IBM's Skills Builder Network. Learning how some simple mathematical operations could be used to make predictions and discover patterns sparked my curiosity, leading me to explore the exciting world of AI. Over the years, I’ve gained hands-on experience in building scalable AI solutions, fine-tuning models, and extracting meaningful insights from complex datasets. I'm driven by a desire to apply these skills to solve real-world problems and make a meaningful impact through AI.

Read more

Magdy Hafez

Software Developer

I really like React

Read more

Brian Ho

UX/Product Designer

UX Designer Intern at IBM | Master of Information in User Experience Design (UXD) Student | UX Research | Design with Agility, Inclusivity, and Accessibility

Read more