IvorySoft portfolio case: Toxicity. Social Media Content Moderation Extension

Social

Browser extension for analyzing social media behavior: developing an extension to understand how people interact on social networks.

Social media often feel like wading through an overwhelming chatter. There’s a lot of noise, blathering, negative discussions, and all sorts of other things. To cut through it, Jigsaw got in touch with the IvorySoft team to craft a browser extension called Toxicity.

About the project

Toxicity is a browser extension that keeps online conversations healthy and respectful. It evaluates comments in real time, scoring them based on emotional attributes, which measures the likelihood of a comment being rude or disrespectful. Users can configure the aspects they want to monitor, and Toxicity will do the rest, providing scores whenever new or updated content appears.

SCOPE OF WORK

For the Toxicity project, our goal was to create a tool that analyzes and moderates user-generated content across various social media platforms. We focused on building the extension from scratch, ensuring it could interact with content and provide valuable insights.

technologies used:

TypeScript

Service Workers

integrations:

Firebase

task 1

Develop backend to support all functionalities of the app.

task 2

Develop a system to analyze and filter toxic posts and comments on social media.

Understanding how Toxicity works

When someone posts a comment, it’s automatically saved in a Firestore collection. The Toxicity extension detects this new comment and reads it. The comment is then sent to the Perspective API for toxicity analysis. The results of this analysis are written back into the Firestore document, making it easy to access and review.

Unpacking how we crafted the Toxicity extension

By combining some serious tech and our expertise, we’ve developed the Toxicity extension to keep your social media browsing sharp and efficient. Here’s a quick look at the tools and tricks the IvorySoft team used.

Dynamic DOM interaction

IvorySoft developers have crafted scripts that efficiently work with the DOM trees of major social networks like Facebook, Twitter X, and Reddit. These scripts are dynamic, ensuring the extension easily adapts to any structural updates on these platforms.

Optimized data collection

We’ve optimized the process of collecting data from social networks to ensure it does not affect user productivity. Service Workers run in the background, handling heavy tasks without impacting the front-end performance.

Unified code base

Our team implemented a flexible and adaptable search algorithm that allows the same code base to function across different social networks, ensuring consistency and maintainability.

Fault-tolerant design

We’ve implemented fault-tolerant design patterns to ensure operational stability. This approach prevents the system from failing in case of an error, maintaining continuous service without interruption.

Logging and monitoring

All operations and incidents are logged into Firebase, which provides the Toxicity extension with real-time monitoring capabilities and aids in swift troubleshooting and analytics.

Parsing algorithm optimization

The parsing algorithms have been fine-tuned using Service Workers to enhance performance. This optimization allows faster processing and a smoother user experience without bogging down the browser.

Summary

We know social media platforms are always changing, with new updates and design tweaks popping up all the time. That’s why we made sure Toxicity is up to the challenge. As a result, our adaptive algorithms adjust to various modifications, keeping the extension running smoothly and users happy.

Powerwatts

Emulait

Looking to create a perfect solution?

Contacts

Links

Portfolio Services Careers Tech Insights