We got inspired by an article * from Plan international Canada that opened our eyes on the Cyberbullying problem. This problem continues to be a serious issue affecting our social life and the mental health of teens, young adults and especially girls. So, we thought it would be a great idea if there was an extension that reminds users to be mindful and reflective when they are about to send an insulting comment/message.

What it does

Our project"Bully-in-spect" is a google chrome extension that gets what the user is currently typing, analyzes the text, and detects if there is any hate or offensive speech.

  • If so, a pop-up box will appear to warn the user to be careful and to take a moment to reflect before sending the message.
  • Another feature is, to analyze the web pages and detects any hate and offensive language. Then it would give the result in a pop-up box!

How we built it/how scalable is it?

We built it with love and programming languages!

  • It is easily scalable to different platforms and browsers as we are using Javascript with the React Framework which is cross-platform widely used everywhere.
  • We used content scripts to trigger page scans for sentiment analysis after each webpage loads.
  • We had background scripts to interact with the Google Cloud Natural Language Processing API to fetch the sentiment scores for each sentence.
  • The popup UI was done with CSS.

Challenges we ran into

With a new project, new challenges come! Some of the challenges we encountered during the Hackathon were

  • How to create a Google Extension
  • Connect Extension with Google Cloud; authenticate
  • Accuracy issue with API
  • Collection of User's Data
  • Handling React Hooks State

Accomplishments that we're proud of

  • I teamed-up with total strange participants and they end up being amazing and creative people!
  • I am proud to keep learning about many new things during the Hackathon.
  • Able to build a functional extension!

How We used Google Cloud

  • Initially we used AutoML to train the Model of Sentiment ,but we didn't get the needed accuracy.
  • Then we used Google Cloud NLP Sentiment API to get the analysis done.
  • We used HTTP POST requests to call the NLP API and get our results.

What we learned

  • We learned to create a Google Chrome Extension
  • We learned about the latest technologies from Daniel and Aniket. Such as: Google Cloud, API, Authentications, and React.
  • It was great working with team members from different timezones and trying to create such a project in 24 Hour Time-Limit

What's next for Bully-in-spect

We would like to go bigger and implement this extension to other larger browser platforms such as safari and firefox and social media platforms (where bullying incidents occur more readily). We would love to create it as an Android-App with extra features:

  • Parental restrictions
  • Highlights the offensive words
  • Show words suggestions for the insulting comments
+ 3 more
Share this project: