Redefining cyberbullying using empathy and inclusivity one cookie at a time.
old beginnings
In 1844, the first electronic communication was sent by a telegraph machine. Tap, long tap, tap, it went, like a rhythmic musical pattern. Little did we know that it was the Big Bang of communication using digital technology.utopia
More than a century later, the urge to enable a computer to communicate with the next birthed the creation of the Internet, which brought exciting new frontiers and landscapes. When personal computers were popularized a decade after, the scenery was set to catapult creativity. Emailing, blogging, online chatting, and bulletin board systems bloomed into a garden for self-expression and creating online communities. Slicing into the 21st century, social media networks had a slow crack in the ice in the human population at first. But with rapid technological progression, a field of social media platforms came to fruition.entropy
Where are we now?
Social media has heated up problems such as algorithmic biases and cyberbullying. It has drawn deep cracks in valuable, human connections. There is now a drought of empathy, as a 1080 x 720 rectangular screen camouflages bots and human emotion. Not to mention, there are floods of bot attacks, riddled with targeted advertisements and data breaches.
Can technology solve the mental health problems it creates? Can we save social media from entropy?
Affective Cookies attempts to tackle this by leveraging the power of natural language processing to create a mental health feed. Empathy is used to generate a custom mental health resource card. Words can be used to form an inclusive message to combat cyberbullying. All the while, cookie points can be earned from positive tasks. The vision for the platform is to provide a space where adolescents can engage in self-care activities while enjoying healthy social media use.
Behind the curtains, the site uses Next.js, React.js, and Tailwind CSS for the front end. Airtable serves as the back end, fetched using SWR and Axios. Sentiment analysis was pretrained from AFINN, designed for Node.js with an 86% accuracy. The inclusive text completion was created using OpenAI. The website was deployed on Vercel.
Made with ❤ by Audrey Wang.