As individuals who spend a lot of time on the internet, the sight of toxicity on the web is nothing new to us. But that does not mean we cant do anything about it. We were hopeful to find a possible solution to this problem, and help clean up the internet
What it does
Our program will take a statement, analyze it for its sentiment (positive, negative and neutral) and accounts for it in out profile of that user on that site or service. We then monitor problem individuals for excessive toxicity and can alert the user or the admins of the chat service to the behavior of the profile, and allow them to take action against toxic individuals without the harshness of a robot or the time consumption of having a human do the work.
How we built it
We build our Interface using a slew of Amazon Web Services. All of our code runs on cloud, on a lambda, that will run only when called and analyze the sentiment of the sentence. This gets passed throughout Amazon's Cloud-front between lambdas to ultimately end in out Amazon DynamoBD. This all written in python and JSON objects
Challenges we ran into
Having no prior experience with AWS, it was a challenge to learn and adapt to this new development style, we had to improvise at some key moments to keep on track, but ultimately we managed to overcome the challenges we faced.
Accomplishments that we're proud of
We get a hold of a new working environment fairly quickly, having never worked with cloud computing before we also managed to display our data in an interesting manner using HighCharts.
What we learned
We learned alot about what AWS can do, and there is even more that we didn't use. We also got some experience with virtual databases and cloud computing, as well as micro services.
What's next for Mean Clean
Scaling the project to be accessible to more end users so they can keep track of people who may be bringing down their communities