Calibration using knowledge graph attributes in recommender systems

O-Chol Kwon; Mingxin Gan

doi:10.1117/12.3004143

11 October 2023 Calibration using knowledge graph attributes in recommender systems

O-Chol Kwon, Mingxin Gan

Proceedings Volume 12800, Sixth International Conference on Computer Information Science and Application Technology (CISAT 2023); 1280059 (2023) https://doi.org/10.1117/12.3004143
Event: 6th International Conference on Computer Information Science and Application Technology (CISAT 2023), 2023, Hangzhou, China

Abstract

Calibration is currently becoming a topic in machine learning. In particular, there have been many attempts to realize fair recommendation through the calibration in recommender systems. However, current research on the calibration in recommender systems has been realized by using genres in the film domain. Few studies have been conducted to generalize the calibration method to other recommendation domains. In this paper, we extend the calibration to other recommendation domains using the attributes of knowledge graphs, which are globalized and standardized knowledge source. We conducted extensive experiments using the weighted matrix factorization algorithm and the graph networks of Amazon and Yelp2018 datasets. The results show that calibration based on knowledge graph attributes improves the accuracy and diversity of the recommendation system and is applicable to all recommendation domains. This method is based on the re-ranking method and can be used with all recommended algorithms.

(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

O-Chol Kwon and Mingxin Gan "Calibration using knowledge graph attributes in recommender systems", Proc. SPIE 12800, Sixth International Conference on Computer Information Science and Application Technology (CISAT 2023), 1280059 (11 October 2023); https://doi.org/10.1117/12.3004143

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available