Universal Document Recommender System

Dylan Vincent


Supervised by Stuart M Allen; Moderated by Stefano Zappala

The idea is to create a system that can recommend documents based on a given document (string) input. The system will use an internal storage of retrievable documents initially, but could later point to a few selected websites and search their databases for documents with similarities to the input document. I will be looking at algorithms for Vector Space Modeling as well as other possible Natural Language Processing methods and comparing them against each other via an evaluation function. This system will be integrated into a website application and hosted via a Raspberry Pi.

Initial Plan (06/02/2021) [Zip Archive]

Final Report (14/05/2021) [Zip Archive]

Publication Form