https://journal.ubaya.ac.id/index.php/saintek/issue/feedKeluwih: Jurnal Sains dan Teknologi2024-10-25T17:02:26+00:00Singgih Sugiartosinggih_s@staff.ubaya.ac.idOpen Journal Systems<p align="justify"><strong>Keluwih: Jurnal Sain</strong><strong>s</strong><strong> dan Teknologi </strong>is an online, open access, and peer-reviewed journal. JST publishes two issues per year: in February (covering February-July) and in August (covering August-January). This journal is to provide a forum for the sharing, dissemination, and discussion of original research, case studies, and critical reviews in the fields of science and technology including biotechnology. This focus and scopes include, but are not limited to subjects in industrial engineering, informatics, electrical engineering, manufacture, environmental issues, renewable energi, chemistry and chemical engineering, product design & management, and fashion design & lifestyle products.</p> <p align="justify"><strong>eISSN: </strong>2721-2432</p>https://journal.ubaya.ac.id/index.php/saintek/article/view/6498Sentiment Analysis for Sumber Gempong Rice Field-Based Tourism Destination using Long Short-Term Memory 2024-10-25T17:02:26+00:00Njoto Benarkahbenarkah@staff.ubaya.ac.idVincentius Riandaru Prasetyovincent@staff.ubaya.ac.idJehuda Rivaldo Soetiyonojehuda1121@gmail.com<p style="text-align: justify;"><strong><em>Abstract</em></strong>—<em>Sumber Gempong is a rice field-based tourist destination located in Ketapanrame Village, Trawas District, Mojokerto Regency, East Java Province. It is managed by a village-owned company (BUMDesa Mutiara Welirang). BUMDesa evaluates tourist satisfaction manually by reviewing online comments and it consumes time and labor works. Data used in this research automatically collected from Google Maps Review. Long Short-Term Memory (LSTM) method analyze data of two sentiment labels, positive or negative, based on four categories: facilities, services, culinary, and attractions. The collected dataset has 674 comments consist of 420 positive sentiments and 254 negative sentiments with 320 facilities, 61 services, 125 culinary, and 192 attractions comments. Five LSTM models were trained on each of four categories and an overall category. The trained models of overall, facilities, services, culinary, and attractions categories achieved, respectively, 91.2%, 86.8%, 94.1%, 89.7%, and 95.6% of accuracies. The average result accuracy is 91.48%. The manager of BUMDesa Mutiara Welirang satisfied with the results of the system and the sentiment results can be used as evaluation material for Sumber Gempong.</em></p> <p style="text-align: justify;"><strong><em>Keyword</em><em>s:</em></strong> <em>sentiment anaylsis, LSTM, deep learning, social media, tourism</em></p> <p style="text-align: justify;"><em> </em></p> <p style="text-align: justify;"><strong>Abstrak</strong>—Wisata Sawah Sumber Gempong berada di Desa Ketapanrame, Kecamatan Trawas, Kabupaten Mojokerto dan merupakan tempat wisata alam yang dikelola oleh BUMDesa Mutiara Welirang. Evaluasi terhadap tempat wisata ini dilakukan dengan membaca secara manual ulasan-ulasan yang ditulis di media sosial dan pengamatan pribadi. Banyaknya jumlah ulasan yang ada menjadi kendala dalam melakukan evaluasi karena membutuhkan waktu yang cukup lama. Penelitian ini mengambil data ulasan secara otomatis dari media sosial yang diberi label positif atau negatif berdasarkan empat kategori, yaitu fasilitas, pelayanan, kuliner, dan wahana. Metode <em>Long Short-Term Memory</em> (LSTM) dipakai sebagai alat untuk melakukan analisis sentimen. Pengambilan data secara otomatis mendapatkan 674 ulasan yang dibagi menjadi 420 ulasan positif dan 254 ulasan negatif, dengan 320 ulasan fasilitas, 61 ulasan pelayanan, 125 ulasan kuliner , dan 192 ulasan wahana. Lima buah model dilatih berdasar tiap kategorinya dan kategori secara keseluruhan. Model yang telah dilatih mendapatkan nilai akurasi sebesar 91,2%, 86,8%, 94,1%, 89,7%, dan 95,6% berturut-turut untuk keseluruhan kategori, kategori fasilitas, layanan, kuliner, dan wahana. Rata-rata akurasi mencapai 91,48%. Hasil dari sistem telah diujicobakan kepada manajer BUMDesa Mutiara Welirang dan bisa dipakai sebagai bahan evalusi untuk peningkatan kualitas di Sumber Gempong.</p> <p style="text-align: justify;"><strong>Kata kunci:</strong><em> analisis sentimen, LSTM, deep learning, media sosial, wisata</em></p>2024-10-31T00:00:00+00:00Copyright (c) 2024 Njoto Benarkah, Vincentius Riandaru Prasetyo, Jehuda Rivaldo Soetiyono