Take10

Inspiration

Mental health and stress management are some of the most important issues in life in 2023, and oftentimes, the tasks we must complete are the source of such issues. Being able to identify and organize tasks based on personal stress levels improves productivity and mental health.

What it does

Take10 lists tasks in a Kanban board based on completion level, and each task includes tags such as task duration, due date, and class. Tasks can be added manually by the user or by uploading a file that would be parsed for assignment information and automatically added. When the user adds tasks manually, an NLP model identifies the task name, due date, and class the task is for. Users can then begin a work session, which utilizes a 01 knapsack problem to identify which tasks should be done during the session and utilizes an image-to-sentiment model to track the stress levels of the user. Whenever the stress levels are identified to be too high, there will be a pop-up telling the user to Take10, or take a ten-minute break before continuing. At the end of the session, users can see a graph displaying their stress level throughout the session.

How we built it

We created a Kanban board with a calming color scheme using next.jss,css. We began working on the UI with a button to add tasks and a logo. We used Redis as a database. We created an NLP Model that would identify the tags of each task and assign a duration and functionality to extract assignments and tasks from any uploaded file. We continued to create an image-to-sentiment model using Python that is continuously training. We then created the necessary UI for users to begin a work session and the relevant models to help the user visualize their stress levels throughout the session. The session would identify which tasks the user should work on with a 01 knapsack algorithm. To identify stress levels during the session, we used a webcam to track stress levels and included a warning that would let users know when their stress levels are too high, prompting them to take a break when this happens.

Challenges we ran into

There was a compatibility issue with the image-to-sentiment model as we could not obtain the data that we needed in the format that we required. Figuring out which algorithm to use for task selection during a session and ensuring that it was an accurate algorithm was a challenge we overcame.

Accomplishments that we're proud of

We were very proud of connecting the facial recognition software in a way that allowed us to get all the data in the format we wanted.

What we learned

We learned how to use sockets and a custom image-to-sentiment model tailored to our needs in order to calculate stress.

What's next for Take10

In the future Take10 will be applicable to more than just students. While the current document parsing can parse any document beyond syllabi, the NLP model is tailored to identify classes and, in the future, will get the ability to identify more categories.