PERFORMANCE ANALYSIS OF A HYBRID RECOMMENDER SYSTEM

Uzair Sultan, Hajra Khan, Yasir Saleem Afridi, Mian Ibad Ali Shah, Muniba Ashfaq, Affera Sultan
Additional contact information
Uzair Sultan, Hajra Khan, Yasir Saleem Afridi, Mian Ibad Ali Shah, Muniba Ashfaq, Affera Sultan: Department of Computer Systems Engineering, UET, Peshawar, Pakistan

International Journal of Innovations in Science & Technology, 2024, vol. 6, issue 5, 257-265

Abstract: In the prevailing information age, human confrontation with extensive information makes it difficult to segregate the relevant content on the basis of choices and priorities. This gives rise to the need for effective recommendation systems that can be incorporated into distinct and diversified domains such as e-commerce, social media, and news media websites and applications. By giving suggestions, these recommender systems efficiently reduce huge information spaces and direct the users toward the items that best match their requirements and preferences. Hence, they play an important role in filtering out the relevant user-specific information. Based on the working principle, recommender systems can be classified into Content-Based Systems, Collaborative Filtering Systems, or P opularity-Based Systems. However, to cope with the problems of cold-start and plasticity that are associated with standalone recommender systems, hybrid recommendation systems are being introduced. This research is therefore focused on the development of a Weighted Hybrid Model that combines the scores of the three standalone recommender models in a linear fashion. The performance of the proposed hybrid model is tested against all three standalone models on an online News dataset. Using a Top-N accuracy metric, it is found that the accuracy of the weighted hybrid model is higher than the standalone Content-Based, Collaborative, and Popularity-Based models against the same dataset. An efficiency of 90% for the Hybrid model was achieved compared to the best-performing standalone model having an efficiency of 53%.

Keywords: Machine Learning; Data Mining; Recommender System; Collaborative Filtering; Hybridization Techniques; Evaluation Metrics; Mean Absolute Error (MAE); Cross-Validation; Training Data (search for similar items in EconPapers)
Date: 2024
References: Add references at CitEc
Citations:

Downloads: (external link)
https://journal.50sea.com/index.php/IJIST/article/view/807/1380 (application/pdf)
https://journal.50sea.com/index.php/IJIST/article/view/807 (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:abq:ijist1:v:6:y:2024:i:5:p:257-265

Access Statistics for this article

International Journal of Innovations in Science & Technology is currently edited by Prof. Dr. Syed Amer Mahmood

More articles in International Journal of Innovations in Science & Technology from 50sea
Bibliographic data for series maintained by Iqra Nazeer ().