ML & IoP based roommate search/compatibility prediction web-app

Finding a good roommate is a tough challenge faced by a major number of college students and working professionals. Poor roommate matching can lead to conflicts, and many other problems, affecting both mental well-being as well as lose of productivity. To solve this problem, the following project is being developed which leverages web app functionalities to enhance roommate-matching process. The system collects user preferences through various interactive web interfaces and applies different clustering algorithms to determine compatibility and shows other users with almost similar preferences and matching compatibility. In the first stage, user data is pre-processed and feature selection is performed using TF-IDF and correlation-based filtering. The second stage utilizes clustering techniques like K-Mean, Database scan etc. to group similar users.