Online harassment is never cool but some people still don't get the memo. Why let this still be issue when we can do something to control it? We decide to do our part by actively looking for the harassers and flagging them as inappropriate or reporting them on social media websites.

What it does

Bully Blocker is a web application tool that allows users to submit a url which shows public comments on any website (i.e. Youtube, Facebook, Devpost) and detects harmful comments posted by users. The web app parses the words used by the commenter and identifies the potential harassers online. After that, Bully Blocker gives the user the option to report/flag the harmful commenter by highlighting the comment.

How I built it

We used python and node.js coupled with a sentimental analysis tool to detect harmful words in a comment's collection of words. Using a parsing tool, the words are filtered according to the sentiment analysis tool to determine if it holds and 'harassing' value. If it 'scores' poorly, the comment gets identified as online harassment and the user gets to choose to report or flag the commenter.

Challenges I ran into

We have issues with implementing the sentiment analysis tool in our application as the API was not truly intuitive or easy to understand. Nevertheless, we tried the best we can and applied Node.js and python in our backend to execute the application.

Accomplishments that I'm proud of

Being able to learn multiple APIs and how to use them in a short period of time is one of the truly greatest accomplishments our team could have done. We were barely beginners in web development and we had to learn JavaScript, CSS, Bootstrap, Node.js and JavaScript as fast as we could and determine the best way to implement our new-found knowledge.

What I learned

That we can do whatever we can as long as we put our mind into it. Online harassment is, unfortunately, still a major issue out there and having the ability to make an impact, no matter how small, is something we're grateful to be able to do.

What's next for Bully Blocker

We aim to enable the web app work across all platforms. That is, we don't want to exclude to just social media websites, but also news articles and blog posts. We need to be able to parse any interactive website and pick up harmful keywords. Eventually, we want to evolve so that each website has a scoring of frequency of online harassment. Doing so would require a better understanding of web scraping and parsing tools as wells as a thorough understanding of sentiment analysis.

Share this project: