Abstract
The recommender system is a knowledge-based filtering system that predicts the users' rating and preference for what they might desire. Simultaneously, the neighborhood method is a promising approach to perform predictions, resulting in a high accuracy based on the common items. This method, furthermore, could affect the resulting accuracy value because when each user provides limited data and sparsity, the accuracy of value might be narrow down as a consequence. In this research, we use the Swarm Intelligent (SI) technique in the recommender system to overcome this problem, whereby SI will train each feature to optimal weight. This technique's main objective is to form better groups of similar users and improve recommendations' accuracy. The intelligent swarm technique used to compare its accuracy to help provide recommendations is the Firefly and Bat Algorithm. The results show that the Firefly Algorithm has slightly better performance than the Bat Algorithm, with a difference in the mean absolute error of 0.02013333. The significance test using the independent t-test method states that no statistically significant difference between Bat and Firefly algorithm.
Highlights
The recommender system is research that is most popular in the business world
This study indicates that Bat Algorithm (BA) has a 6.9% better quality than ABC, obtained a lower root mean squared error (RMSE) score than ABC, and higher precision, recall, and F1 score
To measure the performance of BA, the writer compared it with Particle Swarm Optimization (PSO), where BA obtained better results, both in terms of measuring Mean Squared Error (MAE) as much as 3.84% better and BA reaching 85.54% compared to PSO of 85.54% compared to PSO of 81.85%
Summary
The recommender system is research that is most popular in the business world. The recommender system is developed on several algorithms to find the best pattern from data and provide a recommendation by filtering a preference-based on the user's interests or needs [1], weighted update [2]. In Content-based Filtering, the most defining features are used to model items and users. Neighborhood-based methods provide a study about the relationships between items or between users. Neighborhood-based methods make predictions based on common items that help the accuracy value when each user provides limited data, and the spread, the resulting accuracy value will be small-scale [9]. The addition of the SI uses the recommender technique to learn the optimal weight of each feature. It can form better groups of similar users and improve recommendations. This research tried to combine IS techniques (BA and FA) with the neighborhood-based method to solve the sparsity problem and improve the accuracy of the recommendation system. At the end of the experiment, MAE and RMSE values will be obtained, which show the comparison of the accuracy of the BA and FA in providing recommendations
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
More From: International Journal of Advanced Computer Science and Applications
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.