A Recommendation System for an Open Archive

This poster is part of the Open Repositories 2021 Poster Session which takes place in the week of June 7-10. We encourage you to ask questions and engage in discussion on this poster by using the comments feature. Authors will respond to comments during this week.

Authors:

Gulce Bal Bozkurt and Gozde Boztepe Karatas

Poster description:

In this study, we developed a recommender system for OpenMETU which is the open archive of Middle East Technical University. Our system recommends items by using a content-based approach. In the content-based approach, the properties of items are vital due to the recommendations are based on them. Our recommendation system is based on the author, abstract, title, and subject similarity. To calculate these similarities, first of all, we extract features of each item by using natural language processing algorithms such as TF-IDF and Universal sentence encoder. We use cosine distance to measure the similarity between two items. Later, we give different weights to each of the similarities to calculate the overall score. Our system recommends the most similar 5 items to the visitor who visits an item in our archive.

(Enter the full screen by opening the options menu)

About the authors:

Skip to content