alisalevin / mentor-match Goto Github PK

Mentor Match
{Currently in progress}

This project is a matching algorithm for a Mentorship Program initiative that I'm spearheading for the national Goldwater Scholar Community. The purpose of the algorithm is to best match mentors (older, more experienced Goldwater Scholars) and mentees (undergraduate Goldwater Scholars) based on their STEM fields of study, research backgrounds, career goals, interests, the areas they are seeking to give/receive guidance in, etc. 

In its finished form, the algorithm will use a combination of the FlashText keyword processing library for keyword matching and a basic decision tree logic for numeric factors calculate a matrix of match scores. I'll then use Monte Carlo Tree Search to play out different matching simulations in order to ensure reasonable matches for all 1300+ scholars, as opposed to just a few perfect matches for the first set of mentees/mentors that are processed, and much worse matches for the rest.  

This algorithm will make our program scalable as well as increase matching efficiency and effectiveness.