ArXiv Recommender System
Michael Geis, Ethan Farber, Jenia Rousseva, Jeeuhn Kim, Anirban Sharma
Design a system to recommend a user arXiv articles most relevant to their research interests. This will be based on an initial input of 'liked/disliked' articles, with the capacity to learn from additional user preference input. One goal is to create something with the ability to handle abrupt changes in interest. (E.g. after finishing a project on topic X, the user may want to lower the priority on articles related to X.)
This project will involve creating a dataset.