The Current State of Linked Data-based Recommender Systems

Linked Open Data (LOD) is a key Semantic Web technology that manages, reuses, shares, exchanges, and interoperates knowledge and information spaces from various knowledge domains. The use of LOD and its benefits in recommender systems have emerged a lot throughout the past few years. Many approaches...

Full description

Saved in:
Bibliographic Details
Published in2021 2nd Information Technology To Enhance e-learning and Other Application (IT-ELA) pp. 154 - 160
Main Authors Mahdi, Ahmed Mounaf, Hadi, Asaad Sabah
Format Conference Proceeding
LanguageEnglish
Published IEEE 28.12.2021
Subjects
Complexity theory
DBpedia
Electronic learning
Knowledge engineering
Linked data
Linked Open Data
RDF
Recommender System
Semantic Web
Task analysis
Online AccessGet full text

Cover

More Information
Summary:Linked Open Data (LOD) is a key Semantic Web technology that manages, reuses, shares, exchanges, and interoperates knowledge and information spaces from various knowledge domains. The use of LOD and its benefits in recommender systems have emerged a lot throughout the past few years. Many approaches made use of LOD to provide a recommendation to the user to help them find relevant data close to their needs out of a massive amount of data. All of them verified that using LOD enhances the recommendation task and produces an accurate recommendation. In this paper, a review is presented on the current state of LOD in the field of recommendation system, with a focus on reviewing the various recommendation techniques used, the problems facing the recommendation processes, the importance of using linked data, and the methods used in previous studies to generate a recommendation based on the Linked Open Data. Despite the recent trend to use linked data to improve recommendations, there are still many challenges, especially if more than one dataset is chosen, as heterogeneity and complexity will arise, aside from the issue of evaluating such systems and comparing them in a way that avoids bias and places each method in its proper comparative position.
DOI:10.1109/IT-ELA52201.2021.9773738