At the recent HackPrinceton hackathon, a team of four students tackled the issue of fake news, by building a Chrome extension for Facebook. The extension, called “FiB”, takes signals from Microsoft Cognitive Services API’s to classify news posts as “Verified” or “Non-Verified”. The student’s FiB app utilizes the following Microsoft Cognitive Services APIs:
- Computer Vision OCR operation is used to understand text from Twitter snapshots
- Computer Vision Adult operation is used to classify potentially NSFW content in images.
- Bing Search API is used to search for the topic of the article, and return the most relevant, trustworthy source for the topic.
The signals above get combined with other signals like website reputation and likelihood of malware and phishing; the resulting “Verified” or “Non-Verified” flag gets displayed to the user as part of their Facebook feed via the Chrome extension. The team also wants to use Text Analytics API in the future to help with keyword extraction from the article.
The team is made up of four students with varying coding experience:
- Nabanita De is a second-year Masters in Computer Science student at the University of Massachusetts Amherst
- Anant Goel is a Freshman at Purdue University
- Mark Craft is a Sophomore at the University of Illinois
- Qinglin Chen is a Sophomore at the University of Illinois
Nabanita, Anant, Mark and Qinglin had 36 hours to complete their hack. They were able to get started on Cognitive Services very fast, by signing up for free keys and reading through the multiple code sample at https://www.microsoft.com/cognitive-services .
The Microsoft Cognitive Services crew is happy to see the students’ initiative and the positive pickup it has had so far. We believe this is a good example of how one can bake AI into their app and provide real value to their users. To further support this initiative, Microsoft has created higher-transaction free keys for Nabanita, Anant, Mark and Qinglin.