Integrasi Data Rating Dan Analisis Sentimen Ulasan Untuk Optimalisasi Sistem Rekomendasi Destinasi Wisata Bali

Ramadhan, Rizqi Fachrizatur (2025) Integrasi Data Rating Dan Analisis Sentimen Ulasan Untuk Optimalisasi Sistem Rekomendasi Destinasi Wisata Bali. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5002211010-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (18MB) | Request a copy

Abstract

Bali merupakan salah satu destinasi wisata utama di Indonesia yang menarik jutaan wisatawan setiap tahunnya. Tingginya arus kunjungan tersebut didorong oleh keberagaman destinasi alam, budaya, dan hiburan yang ditawarkan. Namun, keberagaman ini juga dapat menyulitkan wisatawan dalam menentukan pilihan yang sesuai dengan preferensi mereka, terutama ketika harus menyaring informasi dari berbagai ulasan dan rating yang diberikan oleh pengguna lain. Untuk mengatasi permasalahan tersebut, penelitian ini mengembangkan sistem rekomendasi tempat wisata berbasis metode Neural Collaborative Filtering yang mengintegrasikan data rating dan hasil analisis sentimen dari ulasan wisatawan. Analisis sentimen dilakukan menggunakan model hybrid IndoBERT-LSTM-CNN. Hasil klasifikasi digabungkan dengan data rating melalui skema pembobotan alpha, kemudian digunakan sebagai masukan pada model rekomendasi. Berdasarkan evaluasi menggunakan metrik Mean Absolute Error (MAE) dan Root Mean Square Error (RMSE), performa terbaik diperoleh pada klasifikasi dua kelas dengan nilai alpha = 0.5, menghasilkan MAE sebesar 0.3089 dan RMSE sebesar 0.430. Hasil ini terbukti lebih unggul dibandingkan konfigurasi dengan alpha = 0 yang hanya menggunakan sentimen, maupun alpha = 1 yang hanya menggunakan rating. Dengan hasil tersebut, sistem yang dikembangkan mampu memberikan rekomendasi yang lebih akurat dan personal, serta berpotensi mendukung promosi pariwisata di Bali.
=====================================================================================================================================
Bali is one of the main tourist destinations in Indonesia, attracting millions of visitors each year. This high volume of tourism is driven by the diversity of natural, cultural, and entertainment attractions available. However, this diversity can also make it difficult for tourists to choose destinations that match their preferences, especially when they have to filter information from numerous reviews and ratings provided by other users. To address this issue, this study develops a tourism recommendation system based on the Neural Collaborative Filtering method, which integrates rating data and sentiment analysis results from tourist reviews. Sentiment analysis is performed using a hybrid IndoBERT-LSTM-CNN model. Each classification output is combined with rating data through a weighting scheme denoted by alpha, and then used as input for the recommendation model. Based on evaluations using the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) metrics, the best performance is achieved with the two-class sentiment classification and alpha = 0.5, resulting in an MAE of 0.3089 and an RMSE of 0.430. These results outperform configurations using only sentiment alpha = 0 or only ratings alpha = 1. With these findings, the developed system is capable of providing more accurate and personalized recommendations and has the potential to support tourism promotion in Bali.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Sistem Rekomendasi, Analisis Sentimen, Neural Collaborative Filtering. Recommendation System, Sentiment Analysis, Neural Collaborative Filtering
Subjects:	Q Science > QA Mathematics > QA76.9.I58 Recommender systems (Information filtering)
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Rizqi Fachrizatur Ramadhan
Date Deposited:	01 Aug 2025 02:38
Last Modified:	01 Aug 2025 02:38
URI:	http://repository.its.ac.id/id/eprint/125063

Actions (login required)

View Item