The Student Sentiment Sensor (SSS)

Our Northeastern University senior capstone project

by Jack Carson, Tyler Ceballo, Mark Morton, Dave Pleteau, & John Privitera

The premise

Some context

44% of college students in a 2021 USA national study reported symptoms of anxiety and depression (“College students and Depression,” Mayo Clinic Health System, 07-Sep-2021). Although on-campus mental health resources exist, a limited amount of people may make the time for them and many who need help are reluctant to reach out for various reasons.

Our solution

Leveraging 8MP computer vision (CV) from the FER Python library running on a Raspberry Pi we rapidly—in 3.0s—identified passing students who appear distressed and seeded a React-based touchscreen interaction with them.

We predicted each student's core stressors with a combination of the CV module’s facial emotion (classifying out of 7 total) detection and a Python-constructed decision tree of survey-like questions.

Finally, we provided users with curated support resources by email in order to actively advance student mental health by kickstarting the process.

Getting started with the stack

Prerequisites

• Python3 is installed.
• Node.js is installed (Version < 17.0)
  • Note: Do not install node via a package manager like Debian's apt. Consider using nvm.
  • Note: "sudo apt-get install node npm" may work/be a potential fix
• A Linux-type environment works best.

Setup

git clone <remote_url> TODO: it might be necessary to install several packages with pip.

Node local dependencies

First, navigate to the project root directory.

cd ./frontend/fe-filesystem-api/
npm install

Second, open a new terminal and navigate to the project root directory.

cd ./frontend/fe-main/
npm install

Running the app

First, navigate to the project root directory.

cd ./backend/
# `python3` may be substituted for `python` if necessary.
python3 backend.py

Second, open a new terminal and navigate to the project root directory.

cd ./frontend/fe-filesystem-api/
# `yarn` may be substituted for `npm` if desired.
npm start

Third, open a new terminal and navigate to the project root directory.

cd ./frontend/fe-main/
# `yarn` may be substituted for `npm` if desired.
npm start