Abstract

In the era of big data, many works have proposed recommender systems (RSs) to handle the information overload issue. Among them, very few studies have focused on Arabic content. Recently, the Arabic content on the Internet has remarkably expanded due to the increasing number of Arabic web users. Consequently, it is become necessary to have RSs for such content. To overcome this challenge, we decided to develop a new pipeline allowing to use RSs in an Arabic context. Our main goal is to assess RSs when applied to Arabic content. For this reason, we built four new publicly available large scale Arabic datasets for recommendation purposes. Then, we proposed a relevant scheme for preprocessing the constructed datasets. Finally, we evaluated and analyzed the impact of exploiting Arabic content by two widely used recommendation algorithms from the literature. The experiment results showed that the proposed pipeline ensures promising findings, which may inspire the research community to conduct additional studies in this direction.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call