Open AccessCase-based Mobile Tourism Attractions Recommender System
Author(s) -
Septia Rani,
Dimastyo Muhaimin Arifin,
Sheila Nurul Huda,
Dhomas Hatta Fudholi
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1077/1/012009
Subject(s) - recommender system , computer science , mean absolute percentage error , tourism , visitor pattern , mean absolute error , android (operating system) , matching (statistics) , similarity (geometry) , information retrieval , world wide web , data mining , artificial intelligence , mean squared error , geography , statistics , mathematics , image (mathematics) , archaeology , artificial neural network , programming language , operating system
The variations and the increasing number of tourism attractions give tourists more choices to enrich their travel experience. However, it also presents drawbacks, including confusing tourists when planning a trip, because they have to choose from many tourist attractions options that match their preferences. In this study, we propose an algorithm that can generate recommendations of tourist attractions to the user using a case-based reasoning approach. Case-based recommendation is chosen because it gives the user a more personalized recommendation that they can customize later on within the application. Besides, the implementation of this approach in the tourism domain is still hard to find. As many as 217 tourist attraction objects in Yogyakarta city are gathered as the case study. We label each object with three attributes, namely category, visitor type, and activities. A similarity function called the simple matching coefficient is used to calculate the item’s score based on the user specification. For assessing the proposed algorithm, we implement the model as an Android application. To evaluate the recommendation result, we use mean absolute error calculation (MAE) and mean absolute percentage error (MAPE), resulting in 4.1 MAE and 5% MAPE, which is considered highly accurate. Thus, the recommendation from the system is well suited to the user’s preference.