MentorMatch

What is MentorMatch?

MentorMatch is a service designed to pair mentors to mentees. Each user can create a customizable profile, allowing them to select as many preferred partners as they like. When all of the profiles are complete, MentorMatch will generate a mapping that maximizes the happiness of all parties involved.

How does MentorMatch work?

The designation of mentors and mentees naturally forms the structure of a bipartite graph. We generate our assignments via solving a max-weight bipartite matching problem. Our weights are defined as follows:

• Both partners listed each other as preferred (2 points)
• Exactly one partner listed another as preferred (1 point)
• Neither partner listed each other as preferred (0 points)

Matching can then be formalized as follows. Let $n$ be the number of mentors and mentees and $P_n$ the set of $n \times n$ permutation matrices. We solve:

$$\underset{P \in P_n}{\arg\max} \sum_{i=1}^n \sum_{j=1}^n w_{ij} \cdot P_{ij}$$

where the nonzero terms of $P$ correspond to the ids of the matched mentors and mentees. This is solved efficiently via the user of SciPy's linear_sum_assignment function.

How can I see my matching?

Once all preference data is accumulated, the full matching can be generated via the "Getting Started" page. It can also be downloaded as a separate CSV file, which lists the mentor-mentee pairs as well as the matching score.

Build Instructions

The cloud deployment of MentorMatch can be found here. For users wishing to build locally:

git clone https://github.com/colbyendres/MentorMatch
cd MentorMatch
pip install -r requirements.txt

MentorMatch uses Postegres for the database and requires the SECRET_KEY variable for maintaining sessions. The app is configured to set these config variable from an .env file in the top-level directory

echo "DATABASE_URL=your-postgres-url" >> .env
echo "SECRET_KEY=your-secret-key" >> .env

With the configuration variables set, MentorMatch can be run with flask run.