CASA | Devpost

In the modern world, rapidly expanding technology has been drifting us away from the emotions that we as humans strive to connect with. One must look at the benefits of technology to gain an advantage for ourselves. CASA is designed to use media arts with technology to reflect upon ourselves in a busy life. In a way, this web app can be seen as a digital footprint of the heart.

CASA takes in the user's input of their overall feelings and experiences every day, with their preferred art media and language. This information is used to predict their emotions and suggests to them an art piece to relate to. Overall, it allows them to track their mood and becomes a safe place for the user to share their true feelings - like a home (CASA ;D). This lowers the convince of journaling every day and searching the web for media suggestions.

We used NLP (Natural Language Processing) and the Linear SVC Model to associate a particular emotion like joy, anger, fear, etc with the users' input. Based on the emotion, language, and preferred art form we provide an output from the database. The users previously viewed art forms and the associated emotions can also be seen by them.

The challenges we faced were initially deciding on the project itself, the entire team had the right idea but being on different pages and not implementing a fixed plan took a long time. To satisfy everyone's concern we really focused on the initial planning, whereas technical problems arose while working. Some of them include modifying the database for the NLP and testing its accuracy, fixing the different language outputs, and datasets required to store login information. We are proud to have a web app finished, that accomplishes its mission of connecting people to their everyday lives.

We learned many things like time management, communication is essential, and that things can go wrong in a quick second with a bug in your code, overall this hackathon was fun and we all bonded for the love of our project. Some things that are next for CASA are giving suggestions based on user history, adding a plot graph of the user's mental state, increasing accuracy given more data, and personalizing the emotions further.