Claim Missing Document
Check
Articles

Found 3 Documents
Search

Implementasi Support Vector Machine untuk Analisis Sentimen Aplikasi Deepseek Prilindaputra, Brilian; Putri, Dinda Rima Rachcita; Ulinnuha, Nurissaidah
INTEGER: Journal of Information Technology Vol 10, No 1 (2025): April
Publisher : Fakultas Teknologi Informasi Institut Teknologi Adhi Tama Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31284/j.integer.2024.v10i1.7541

Abstract

Kemunculan DeepSeek, AI canggih yang dikembangkan di China, telah memberikan dampak yang signifikan terhadap lanskap teknologi global. Namun, pengadopsiannya telah mendapat reaksi beragam, dengan beberapa negara memilih untuk memblokir aksesnya karena masalah keamanan data. Penelitian ini bertujuan untuk menganalisis sentimen pengguna terhadap aplikasi DeepSeek di Google Play Store, secara khusus menargetkan ulasan pengguna dari Amerika Serikat. Dengan menggunakan metode klasifikasi Support Vector Machine (SVM), analisis sentimen dilakukan untuk mengkategorikan opini pengguna ke dalam sentimen positif, netral, dan negatif. Dataset yang terdiri dari 10.000 ulasan yang dikumpulkan melalui web scraping, telah dipreproses menggunakan teknik pembersihan teks, pembobotan TF-IDF, dan lemmatization. Model SVM dilatih dan divalidasi menggunakan k-fold cross validation (k-fold = 10), mencapai akurasi terbaik pada parameter C = 100 dan kernel RBF. Hasil evaluasi menunjukkan akurasi rata-rata 90,33%, dengan akurasi puncak 92,20% pada fold 10. Temuan ini menunjukkan polaritas sentimen yang kuat di antara para pengguna. Penelitian ini penyebaran kata dari analisis wordcloud memberikan wawasan bagi para pengembang dan pemangku kepentingan dalam meningkatkan aplikasi AI dengan mengatasi kekhawatiran pengguna dan meningkatkan kepuasan pengguna secara keseluruhan.
Edukasi Pemilahan Sampah Melalui Program Bank Sampah di Desa Maron Wetan Roni, M. Imam Sya; Prilindaputra, Brilian; Mujaddidah, Putri Yasmin Imarotul; Rohmah, Nailur; Mayanti, Ardita Fajria; Jannah, Elly Uzlifatul
Jurnal Kesehatan Lingkungan Mandiri Vol 4 No 1 (2025): Jurnal Kesehatan Lingkungan Mandiri, Volume 4 No.1 November 2025
Publisher : Jurusan Kesehatan Lingkungan Poltekkes Kemenkes Padang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33761/jklm.v4i1.2231

Abstract

Waste management remains a pressing challenge in Maron Wetan Village, Probolinggo Regency, East Java. The situation is exacerbated by the community’s low awareness of waste separation, limited supporting infrastructure, and the influx of waste from outside the area. To address these issues, the Community Service Program (KKN) Group 52 of UIN Sunan Ampel Surabaya applied the Participatory Action Research (PAR) approach, which emphasizes collaboration and active engagement of local residents as key actors rather than passive research subjects. The program commenced with the mapping of 13 neighborhood units (RT) and field data collection, followed by analysis using the Matrix Ranking technique to determine priority interventions based on urgency, efficiency, and impact. The analysis identified three main priorities: low awareness of the long-term environmental impacts of waste (score 9), habitual disposal of waste into rivers (score 8), and inadequate waste management facilities (score 6). In response, two primary interventions were developed: (1) the establishment of a waste bank as an economic incentive scheme, and (2) the installation of educational signboards titled “How Long Does Your Waste Take to Decompose?” to promote environmental consciousness. Positive community responses were particularly evident in RT 2, RT 6, and RT 9, where active participation was observed. These findings highlight that an educational–participatory approach can enhance environmental awareness and encourage behavioral change, although the long-term sustainability of the program remains contingent upon structural support from the village government.
Landslide Modeling with the Savage-Hutter Approach Using the Finite Volume Method Prilindaputra, Brilian; Toyibah, Syifa Nasiratun; Putri, Dinda Rima Rachcita; Novitasari , Dian Candra Rini
International Journal of Mechanical Computational and Manufacturing Research Vol. 14 No. 4 (2026): February: Mechanical Computational And Manufacturing Research
Publisher : Trigin Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/computational.v14i4.285

Abstract

Landslides are one of the most frequent disasters in Indonesia and have a major impact on the environment and society. This study focuses on modeling the dynamics of landslides in Peniraman Hill, West Kalimantan, using the Savage-Hutter (SH) model solved through the finite volume method (FVM) and the Harten-Lax-van Leer flux scheme. (HLL), supported by the Courant–Friedrichs–Lewy (CFL) method to maintain stable conditions. This study aims to apply the model to real conditions and assess the effectiveness of the numerical approach in describing the movement of land masses. Simulations were conducted on Slopes 1 and 3 which are at risk of landslides due to their soil stability, with three variations of the soil friction angle  to see how changes in these parameters affect the flow mechanism and sliding distance. The results show that the soil friction angle  is a factor that influences landslide behavior. Decreasing the value  makes the landslide move faster and cover a wider area in all parts of the topography. The initial maximum velocity of Slope 1 ranges from ~12–17 m/s with a range of around ~18 meters, while on Slope 3 it reaches ~20–27 m/s with a range of up to ~23.5 meters. Slope 3 consistently produces faster movement and longer sliding distance. Overall, the combination of the SH model with the FVM method and the HLL scheme controlled by CFL conditions has proven to be effective, stable, and capable of representing landslide dynamics. The research results can be an important basis for risk analysis and disaster mitigation strategy planning in the environment around Peniraman Hill to establish exclusion zones and design high load-bearing structures in the potential landslide reach area of ~23.5 meters