Implementasi Sistem Rekomendasi Objek Wisata Dengan Metode Hybrid Content-Based, Collaborative Dan Context-Aware Filtering

Lita, Ivana (2025) Implementasi Sistem Rekomendasi Objek Wisata Dengan Metode Hybrid Content-Based, Collaborative Dan Context-Aware Filtering. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5026211058-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Indonesia memiliki potensi pariwisata yang sangat besar dengan keberagaman destinasi yang tersebar di berbagai daerah. Namun, banyaknya pilihan destinasi wisata justru dapat menimbulkan information overload, sehingga menyulitkan bagi wisatawan dalam menentukan pilihan yang sesuai dengan preferensi dan kebutuhannya. Tugas akhir ini bertujuan untuk mengembangkan sistem rekomendasi objek wisata berbasis hybrid filtering yang mengintegrasikan tiga pendekatan utama, yaitu content-based filtering, collaborative filtering, dan context-aware filtering. Sistem ini dirancang untuk menghasilkan rekomendasi yang lebih personal dan relevan berdasarkan profil demografis pengguna serta konteks perjalanan. Pengembangan sistem dilakukan menggunakan dataset "Indonesia Tourism Destination" dari Kaggle yang mencakup 437 objek wisata di lima kota besar Indonesia. Metode hybrid filtering mengintegrasikan content-based filtering berbasis kemiripan profil demografis pengguna, collaborative filtering berdasarkan popularitas kategori dari agregasi rating, dan context-aware filtering yang mempertimbangkan tipe perjalanan pengguna. Sistem menghasilkan pemeringkatan semua 6 kategori wisata dengan masing-masing disertai maksimal tiga objek wisata dengan rating tertinggi dari pengguna dengan profil demografis serupa. Antarmuka pengguna dikembangkan menggunakan framework Streamlit dengan desain yang interaktif dan responsif. Evaluasi usabilitas dilakukan menggunakan instrumen System Usability Scale (SUS) yang melibatkan 35 responden, dan menghasilkan skor 88 yang termasuk dalam kategori “Excellent”. Hasil ini menunjukkan bahwa sistem mampu memberikan pengalaman penggunaan yang baik serta rekomendasi yang sesuai dengan karakteristik dan kebutuhan pengguna. Namun, sistem menghasilkan rekomendasi yang sama untuk pengguna dengan profil demografis yang identik, yang dapat mengurangi variasi eksplorasi wisata.
================================================================================================================================
Indonesia has immense tourism potential, offering a wide variety of destinations across the archipelago. However, the abundance of available options can lead to information overload for tourists, making it difficult to choose destinations that match their preferences and needs. This final project aims to develop a tourism recommender system based on hybrid filtering by integrating three main approaches: content-based filtering, collaborative filtering, and context-aware filtering. The system is designed to provide more personalized and relevant recommendations based on users' demographic profiles and travel contexts. The system was developed using the “Indonesia Tourism Destination” dataset from Kaggle, which includes 437 tourist destinations across five major cities in Indonesia. The hybrid filtering method integrates content-based filtering based on user demographic profile similarity, collaborative filtering based on category popularity from rating aggregation, and context-aware filtering that considers the type of user's trip. The system generates rankings for all 6 tourism categories, with each category displaying up to three tourist destinations with the highest ratings from users with similar demographic profiles. The user interface was developed using the Streamlit framework with an interactive and responsive design. Usability evaluation was conducted using the System Usability Scale (SUS) involving 35 respondents, yielding a score of 88 categorized as “Excellent.” This result indicates that the system provides a satisfying user experience and delivers recommendations aligned with user characteristics and needs. However, the system generates consistent recommendations for users with identical demographic profiles, which may reduce tourism exploration diversity.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Sistem Rekomendasi Pariwisata, Tourism Recommender Systems, Hybrid Filtering, Content-Based Filtering, Collaborative Filtering, Context-Aware Filtering, System Usability Scale
Subjects:	Q Science > QA Mathematics > QA76.9.I58 Recommender systems (Information filtering) T Technology > T Technology (General) > T58.62 Decision support systems
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 57201-(S1) Undergraduate Thesis
Depositing User:	Ivana Lita
Date Deposited:	17 Jul 2025 08:55
Last Modified:	18 Jul 2025 03:20
URI:	http://repository.its.ac.id/id/eprint/119925

Actions (login required)

View Item