Inspiration
I was just going to get some food and it kinda popped up.
What it does
Changes facebook messages sent to you to happy messages if they are detected to be harassment or hurtful.
How I built it
Built with a chrome extension that gets information from the facebook page, sends it to a node.js server that runs IBM Watson, then returns data that determines if the text will be changed into happy text.
Challenges I ran into
I didn't know any of these languages or web dev before this, never touched it. Getting that to work was hard. Hardest part though was IBM Watson, easily. Implementation of it is a nightmare, it's so confusing.
Accomplishments that I'm proud of
Actually making something in a day when I was learning the language at the same time, and finally getting into web dev a bit. Also making my friends laugh when their shit talk turns to fluff.
What I learned
javascript, html, jquery, node.js, how to make a server, html, how to make a chrome extension, how to scrape data, how to send and recieve server requests, how nice it is to have a hackathon five minutes away from your dorm.
What's next for HarassMend
Implementation into other social media sites such as twitter, instagram, etc. Adding to a mobile environment to affect things like snapchat. Adding customization like sliders to affect filter intensity or selecting which sites it would work on. Right now it works on click to demo better but if it's in the real world it should work on end of load of a page (which is simple, just put at content scripts in manifest.json run_at : document_end or whatever). Also I'd like to use IBM Watson's alchemyAPI to take the tone for long conversation to detect if shit talk is playful or sarcastic as that would trigger harassment currently. Oh and maybe implement text to speech so you hear feel good messages,maybe even when you get any message.
Log in or sign up for Devpost to join the conversation.