Our parents/grandparents have been exposed to harmful misinformation online, which has negatively impacted our families. This inspired us to create a tool that not only protects our loved ones but also empowers others to navigate the web more safely and critically.
Bullsift acts as a "firewall" against fake news. It assesses the safety and credibility of websites, alerts users if a site is untrustworthy, and offers a game to explain why the content is flagged, providing users with factual information for a deeper understanding. With this tool, even tech-savvy grandkids will be asking grandparents if the news is real or fake!
We developed a Chrome extension that seamlessly integrates into your browsing experience, offering real-time content analysis for fake news detection. Powered by Cohere's Command R Model, our multi-layer detection system ensures users are equipped with tools to verify the credibility of online information. Each layer of detection is carefully engineered to deliver comprehensive insights, ensuring a smooth and intuitive user experience. The layers include:
- Sensationalism Detection: Used to spot exaggerated phrases within articles.
- Source Credibility Verification: Cross-referencing sources to evaluate trustworthiness.
- Fact-Checking Automation: Comparing claims to trusted news outlets to highlight any discrepancies.
- Fakeness Scoring System: Providing a detailed "fakeness score," based on bias and credibility analysis.
- Defining and grading truthfulness was complex, requiring accurate detection of bias and sensationalism.
- Integrating multiple API calls synchronously for the multi-layer detection system was technically challenging.
- Making the experience fun and engaging, especially for grandparents.
- Developing a highly accurate detection system that flags misinformation in real-time using Cohere’s powerful AI.
- Creating a fun and engaging learning experience for grandparents, encouraging critical thinking and empowering them to spot fake news.
- Designing a user-friendly, interactive, and aesthetically pleasing UI/UX that simplifies complex processes.
- Using Cohere’s LLM models in different ways to achieve the fakeness detection system
- Grinding with 2 hours of sleep!
Our next step is to implement a feature that allows parents to monitor their children's exposure to fake news, offering a safeguard against repeated exposure to harmful misinformation and encouraging healthier online habits.
- Clone this repository to your local machine.
- Visit
chrome://extensionsin your (Google Chrome) browser - Toggle 'Developer mode' on
- Click 'Load Unpacked' and select the cloned directory (make sure it's unzipped if you downloaded it directly from this repo)
- Open a new tab and click the extension's icon to open its popup & see it in action