Ziana Mega Devi, Noor Akhmad Setiawan, Teguh Bharata Adji, Triyanna Widiyaningtyas
One of the social media-based event manager applications that have been developed on the client side is MeetingYuk, and the application on the partner side is known as MerchantYuk. The development of these two applications (MeetingYuk and MerchantYuk) increased the number of user and partner data. Currently, a list of partners is shown to the user without any filtering option against user preferences. The more list of partners provided, the more options for users will increase, thus making users confused about choosing a partner. Based on this problem, it is necessary to develop a recommendation system to help meet users' needs in selecting partners. The recommendation system utilizes a hybrid filtering method, combining collaborative and content-based filtering methods. The collaborative filtering method analyzes the relationship between the ratings given by the MeetingYuk application users, and the content-based filtering method analyzes the similarity of services at each merchant. The two methods were combined using social aperture to get the prediction results. The experiment results show that the hybrid filtering method provides the best results for predicting suitable merchants, with an MAE value of 0.2019 and an RMSE of 0.5161. © 2022 ACM.
Department of Information Technology and Electrical Engineering, Universitas Gadjah Mada, Indonesia; Department of Electrical Engineering, Universitas Negeri, Malang, Indonesia