Claim Missing Document
Check
Articles

Found 2 Documents
Search

EVALUASI SENTIMEN PENGGUNA TERHADAP APLIKASI BANK SAQU DENGAN METODE ALGORITMA NAÏVE BAYES Adhiyasya, Rakha; Simanjuntak, Sry Intan; Bali, Aprida Bertha; Fadholi, Muhammad Farhan
Kohesi: Jurnal Sains dan Teknologi Vol. 7 No. 1 (2025): Kohesi: Jurnal Sains dan Teknologi
Publisher : CV SWA Anugerah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.3785/kohesi.v7i1.11061

Abstract

Studi ini bermaksud untuk mengevaluasi sentimen pengguna mengenai aplikasi Bank Saqu dengan memakai pendekatan komparasi algoritma Naïve Bayes. Evaluasi sentimen merupakan langkah-langkah guna mengenali dan mengklasifikasikan pendapat dan perasaan pengguna ke dalam kategori positif, negatif, atau netral berdasarkan ulasan yang diberikan. Algoritma Naïve Bayes dipilih karena efisiensinya dalam memproses data teks serta kemampuannya memberikan hasil klasifikasi yang kompetitif. Kumpulan data yang dimanfaatkan dalam kajian ini berasal dari tinjauan pemakai aplikasi Bank Saqu yang diperoleh melalui platform digital. Proses penelitian melibatkan tahap praproses data, seperti pembersihan teks, stemming, dan tokenisasi, untuk memastikan kualitas data yang dianalisis. Selanjutnya, dilakukan perbandingan kinerja beberapa bentuk algoritma Naïve Bayes, seperti Multinomial, Bernoulli, dan Gaussian, berdasarkan metrik evaluasi seperti akurasi, presisi, recall, dan F1-score. Temuan riset mengindikasikan bahwa algoritma Naïve Bayes sanggup mengelompokkan opini pemakai dengan tingkat keakuratan yang signifikan. Varian algoritma tertentu menunjukkan kinerja yang lebih unggul dalam menangani dataset yang tidak seimbang. Temuan ini diperkirakan mampu mendukung pembuat perangkat lunak Bank Saqu dalam mengerti pandangan konsumen dan meningkatkan layanan berdasarkan hasil analisis sentimen. This study aims to evaluate user sentiment regarding the Bank Saqu application using a comparative approach of the Naïve Bayes algorithm. Sentiment evaluation involves steps to recognize and classify users' opinions and feelings into categories such as positive, negative, or neutral based on the reviews provided. The Naïve Bayes algorithm was chosen due to its efficiency in processing text data and its ability to deliver competitive classification results. The data used in this study is sourced from reviews of the Bank Saqu application obtained through digital platforms. The research process includes data preprocessing stages, such as text cleaning, stemming, and tokenization, to ensure the quality of the data being analyzed. A comparison of the performance of several Naïve Bayes algorithm variants, such as Multinomial, Bernoulli, and Gaussian, is then conducted based on evaluation metrics like accuracy, precision, recall, and F1-score. The findings indicate that the Naïve Bayes algorithm can classify user opinions with a significant level of accuracy. Certain algorithm variants show superior performance in handling imbalanced datasets. These findings are expected to assist the developers of the Bank Saqu application in understanding customer views and improving services based on sentiment analysis results.
Analisis Kontribusi Pemain Sepak Bola Eropa Berbasis C4.5 Adyatma, Mochammad Eric; Adhiyasya, Rakha; Ryandhika, Rifaldi; Fadholi, Muhammad Farhan; Ajiansyah, Muhammad Rafly; Isak, Yorrel Jensek; Zalogo, Andrianus; Dakhi, Lisbet
Jurnal Ilmu Komputer dan Informatika | E-ISSN : 3063-9026 Vol. 2 No. 3 (2026): Januari - Maret
Publisher : GLOBAL SCIENTS PUBLISHER

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Modern football has evolved into a data-driven industry where statistical analysis is widely used to objectively evaluate player performance. This study aims to classify the level of goal contribution of top European football players using the C4.5 algorithm implemented through RapidMiner. The dataset is derived from player statistics of top European leagues in the 2022–2023 season, with key attributes including Shot on Target Percentage (SoT%), Shot-Creating Actions (SCA), standardized playing time (90s), and short and total passing accuracy. The research methodology consists of data selection, preprocessing, data transformation, data mining, and model evaluation. The C4.5 algorithm is applied using the Gini Index criterion with pruning techniques to prevent overfitting. Model validation is conducted using 10-Fold Cross Validation. The results show that the classification model achieves an accuracy of 90.67%, with SoT% identified as the most influential variable, followed by SCA and playing time. The generated decision tree provides clear and interpretable rules, making it useful as a decision-support tool for evaluating player contributions based on data-driven analysis.