cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
Mesran
Contact Email
mesran.skom.mkom@gmail.com
Phone
+6282370070808
Journal Mail Official
jurnal.bulletincsr@gmail.com
Editorial Address
Jalan sisingamangaraja No 338 Medan, Indonesia
Location
Kota medan,
Sumatera utara
INDONESIA
Bulletin of Computer Science Research
Published by Forum Kerjasama Pendidikan TInggi
ISSN : -     EISSN : 27743659     DOI : -
Core Subject : Science,
Computer Science & IT
Bulletin of Computer Science Research covers the whole spectrum of Computer Science, which includes, but is not limited to : • Artificial Immune Systems, Ant Colonies, and Swarm Intelligence • Bayesian Networks and Probabilistic Reasoning • Biologically Inspired Intelligence • Brain-Computer Interfacing • Business Intelligence • Chaos theory and intelligent control systems • Clustering and Data Analysis • Complex Systems and Applications • Computational Intelligence and Soft Computing • Distributed Intelligent Systems • Database Management and Information Retrieval • Evolutionary computation and DNA/cellular/molecular computing • Expert Systems • Fault detection, Fault analysis, and Diagnostics • Fusion of Neural Networks and Fuzzy Systems • Green and Renewable Energy Systems • Human Interface, Human-Computer Interaction, Human Information Processing • Hybrid and Distributed Algorithms • High-Performance Computing • Information storage, security, integrity, privacy, and trust • Image and Speech Signal Processing • Knowledge-Based Systems, Knowledge Networks • Knowledge discovery and ontology engineering • Machine Learning, Reinforcement Learning • Networked Control Systems • Neural Networks and Applications • Natural Language Processing • Optimization and Decision Making • Pattern Classification, Recognition, speech recognition, and synthesis • Robotic Intelligence • Rough sets and granular computing • Robustness Analysis • Self-Organizing Systems • Social Intelligence • Soft computing in P2P, Grid, Cloud and Internet Computing Technologies • Support Vector Machines • Ubiquitous, grid and high-performance computing • Virtual Reality in Engineering Applications • Web and mobile Intelligence, and Big Data • Cryptography • Model and Simulation • Image Processing
Arjuna Subject : Ilmu Komputer (semua kategori) - Ilmu Komputer (Lain-Lain)
Articles 310 Documents
 27   28   29   30   31 
Perbandingan Kinerja Arsitektur MobileneTV2 dan MobileneTV3 Dalam Klasifikasi Penyakit Retina pada Citra Optical Coherence Tomography (OCT) Menggunakan Optimizer AdamW dan SGD Ricko Andreas Kartono; Nur Rachmat
Bulletin of Computer Science Research Vol. 6 No. 1 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bulletincsr.v6i1.915

Abstract

Retinal diseases are serious visual disorders that can lead to decreased visual function and even blindness. The diagnosis of retinal diseases is generally still performed manually by medical professionals through the examination of Optical Coherence Tomography (OCT) images, a process that requires considerable time, high precision, and is prone to diagnostic errors. Previous studies have mostly employed larger and more complex CNN architectures, with optimization limited to a few commonly used optimizers. This study aims to develop an automatic retinal disease classification model using Convolutional Neural Network (CNN) methods by leveraging the lightweight and efficient MobileNetV2 and MobileNetV3 architectures, enabling faster applications that can be deployed on resource-constrained devices. The architectures evaluated include MobileNetV2, MobileNetV3-Large, and MobileNetV3-Small, along with a comparison of two optimizers, namely AdamW and Stochastic Gradient Descent (SGD). The dataset used consists of 4,000 OCT images divided into four classes: Normal, Choroidal Neovascularization (CNV), Diabetic Macular Edema (DME), and Drusen. The training process was conducted using a transfer learning approach, and model performance was evaluated using accuracy, precision, recall, and F1-score metrics. The experimental results indicate that the combination of the MobileNetV2 architecture with a batch size of 16 and either the AdamW or SGD optimizer achieved the best performance, reaching an accuracy of 85.75%, which is the highest among all tested configurations. These findings highlight the strong potential of lightweight architectures to be developed into fast, accurate, and field-deployable retinal disease diagnostic applications on mobile devices using deep learning.
Sistem Pendukung Keputusan Pemilihan Content Creator TikTok Terbaik Menggunakan Metode Simple Additive Weighting (SAW) dengan Pembobotan Rank Order Centroid (ROC) Jahraini, Ratu Adnin; Sanwani, Sanwani; Sulthon, Besus Maula; Siregar, Syamsimahara; Hutabarat, Pahotton; Mesran, Mesran
Bulletin of Computer Science Research Vol. 6 No. 1 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bulletincsr.v6i1.918

Abstract

In this selection of the best TikTok Content Creator using the Rank Order Centroid (ROC) and Simple Additive Weighting (SAW) methods. ROC and SAW were chosen because they are simple and efficient and can allow proper weighting of criteria and proper ranking. Many factors must be considered in choosing TikTok Content Creators. The selection of the best TikTok content creators is useful for helping audiences assess and select quality content and can add insight and entertainment that is useful for the future. Judging must also be based on predetermined criteria. There are five criteria that are assessed such as creativity, originality, content quality, interaction, ethics. The results of the assessment of each criterion will be normalised and the total weight calculated. The system will sort from the highest value as the first and last rank. With the existence of a decision support system that uses the ROC and SAW methods, it can make it faster and easier to make decisions to choose the best TikTok Content Creator according to predetermined criteria. Ranking results in applying the SAW method shown in table 5 above, the best alternative is the first rank produced, namely A10 with a final acquisition value of 0.956. The second rank is alternative A3 with a final value of 0.935 and the third rank is alternative A1 with a final value of 0.909.
Penerapan Seleksi Fitur Information Gain dan Metode Backpropagation Neural Network Untuk Klasifikasi Atrisi Karyawan Dinyah Fithara; Elvia Budianita; Iis Afrianty; Siska Kurnia Gusti
Bulletin of Computer Science Research Vol. 6 No. 1 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bulletincsr.v6i1.922

Abstract

Employee attrition management is a critical challenge for organizations as it involves costs, time, and the risk of decision-making errors. This problem requires a data-driven business strategy to achieve more accurate predictions of employees who are potentially at risk of termination. This study applies the Information Gain feature selection method and the Backpropagation Neural Network (BPNN) algorithm in the employee attrition classification process with the aim of increasing the accuracy and efficiency of the prediction model. BPNN is chosen due to its simpler architecture, faster training time, and greater stability for small to medium sized datasets.  With the assistance of Information Gain feature selection, BPNN is able to achieve optimal performance without requiring a complex architecture. The dataset used consist of 35 attributes and 1.470 employee records covering various factor such as age, income level, and employment status. The research stages include feature selection based on information gain values with specific thresholds, data partitioning using k-fold cross validation, and model training using BPNN with variations of learning rates and hidden neuron counts. The results show that the combination of Information Gain and BPNN improves classification accuracy compared to models without feature selection, achieving the highest average accuracy of 87.28% when using 25 selected attributes, with a BPNN configuration of learning rate 0.001, 35 hidden neurons, and 50 epochs. The attributes with the highest Information Gain score include JobLevel, OverTime, MaritalStatus, and MonthlyIncome. This study demonstrates that the proposed approach successfully enhances the prediction performance of employee attrition and can serve as a foundation for developing data-driven models that support employee retention efforts.
Analisis Komparasi Metode SAW Dan TOPSIS Dalam Pemilihan Distributor Barang Gudang Sihmawanto, Fahreza Dandy; Wijayatno, Ganef Tri; Pungkasanti, Prind Triajeng
Bulletin of Computer Science Research Vol. 6 No. 1 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bulletincsr.v6i1.925

Abstract

Distributor selection is a strategic decision in supply chain management that affects product availability, operational costs, and customer satisfaction. This study aims to compare and evaluate distributor selection decision-making using two Multi-Criteria Decision Making (MCDM) methods, namely Simple Additive Weighting (SAW) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). Primary data was obtained through structured interviews with three expert practitioners at a distribution company. Five evaluation criteria were used: product quality (weight 0.30), product price (0.20), delivery accuracy (0.20), service response (0.15), and stock availability (0.15). Ten distributor alternatives were assessed using a 1–10 scale and processed independently using SAW normalization and TOPSIS vector normalization to generate rankings from each method. The results show that PT. Mitra Listrik Nusantara ranked first in both methods with a SAW score of 0.9414 and a TOPSIS score of 0.8057. Ranking consistency comparison through Spearman correlation analysis yielded a value of ? = 0.9515, indicating very high agreement between the two methods. This study provides practical contributions to warehouse management in adopting a data-driven approach for measurable and accountable distributor selection.
Penerapan Metode Support Vector Machine Untuk Analisis Sentimen Pada Komentar Bitcoin Di Aplikasi X Yaskur Bearly Fernandes; Elin Haerani; Fadhilah Syafria; Muhammad Fikry; Lola Oktavia
Bulletin of Computer Science Research Vol. 6 No. 1 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bulletincsr.v6i1.928

Abstract

Social media has become a primary medium for users to express opinions, including those related to Bitcoin, whose fluctuating value often triggers diverse public responses. The large volume of unstructured comments makes manual sentiment analysis inefficient, thereby necessitating an automated approach based on machine learning. This study aims to classify positive and negative sentiments in Bitcoin-related comments on the X platform using the Support Vector Machine (SVM) algorithm with Term Frequency–Inverse Document Frequency (TF-IDF) feature weighting. The dataset consists of 1,750 Indonesian-language comments labeled by three annotators. The data were processed through several preprocessing stages, including case folding, text cleaning, tokenization, stopword removal, and stemming. Model evaluation was conducted using four data split ratios, namely 90:10, 80:20, 70:30, and 60:40. The experimental results indicate that the 90:10 ratio achieved the best performance, with an accuracy of 72.57%, precision of 0.75, recall of 0.73, and an F1-score of 0.67. The SVM model demonstrates strong performance in identifying positive sentiments; however, it is less effective in detecting negative sentiments due to class imbalance in the dataset. As an additional experiment, testing was performed using a balanced dataset obtained through an undersampling process and several SVM kernel types for comparison. The results show that using a balanced dataset leads to more evenly distributed classification performance across sentiment classes, while the linear kernel provides the most stable performance compared to other kernels. Overall, SVM with TF-IDF weighting proves to be an effective approach for sentiment analysis of Bitcoin-related comments on social media.
Perbandingan Naïve Bayes dan SVM untuk Analisis Sentimen Ulasan Kompas.id pada Data Tidak Seimbang Muhammad Ardana; Rini Mayasari; Iqbal Maulana
Bulletin of Computer Science Research Vol. 6 No. 1 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bulletincsr.v6i1.929

Abstract

The rapid advancement of digital technology and the increasing use of mobile devices have driven the widespread adoption of digital news applications, including Kompas.id. User reviews on the Google Play Store represent an important data source for understanding user satisfaction and emerging issues; however, the large volume of reviews makes manual analysis inefficient. Therefore, this study aims to compare the performance of Naïve Bayes and Support Vector Machine (SVM) algorithms in classifying Kompas.id user reviews into positive, neutral, and negative sentiments. The research employs the Knowledge Discovery in Databases (KDD) framework, which includes web scraping, text preprocessing, lexicon-based sentiment labeling, feature extraction using Term Frequency–Inverse Document Frequency (TF-IDF), and classification and evaluation stages. The dataset consists of 1,023 cleaned reviews after data preprocessing. Model performance is evaluated using a confusion matrix with accuracy, precision, recall, and F1-score metrics. The results indicate that Naïve Bayes achieves an accuracy of 72%, while SVM outperforms it with an accuracy of 80%, reflecting its stronger ability to handle high-dimensional and sparse textual feature spaces. Word cloud visualization reveals that positive sentiments are mainly associated with content quality, whereas negative sentiments are dominated by subscription-related issues and technical problems. Based on these findings, SVM is recommended as a more effective algorithm for sentiment analysis of digital news application reviews.
Implementation of the LightGBM–CatBoost Ensemble Method for Obesity Risk Classification in Productive Age Harefa, Kecitaan; Priambodo, Joko
Bulletin of Computer Science Research Vol. 6 No. 1 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bulletincsr.v6i1.930

Abstract

Obesity is a health problem that continues to increase among individuals of productive age and has the potential to reduce quality of life and work productivity. One of the main challenges in obesity risk assessment is the limitation of conventional methods in accurately identifying obesity risk when dealing with complex, multidimensional data that include both numerical and categorical variables. Therefore, an artificial intelligence–based approach is required to provide a more accurate and stable obesity risk classification. This study aims to implement and evaluate a LightGBM–CatBoost ensemble method for obesity risk classification with a focus on the productive age population. The dataset used in this study was obtained from the Kaggle platform and consisted of 2,111 individual records containing physical attributes, eating habits, physical activity, and lifestyle factors. Although the dataset is synthetic and balanced, the included attributes and age-related variables are representative of individuals within the productive age range, making it suitable for modeling obesity risk in this demographic context. The research stages include data preprocessing, separate training of the LightGBM and CatBoost models, model integration using a probability averaging ensemble technique, and performance evaluation using accuracy, precision, recall, and F1-score metrics. The results indicate that both LightGBM and CatBoost achieved accuracy levels above 95%, while the ensemble model demonstrated superior performance with an accuracy of 96.69% and more balanced evaluation metrics across all obesity risk classes. These findings confirm that the ensemble approach improves classification stability and accuracy compared to single models. Therefore, the LightGBM–CatBoost ensemble method is effective for obesity risk classification and has the potential to be further developed as a decision support system in the health sector.
Klasifikasi Sentimen Bitcoin Terhadap Komentar Di Aplikasi X Menggunakan Metode Decision Tree C4.5 Indrizal, Habibi Putra; Syafria, Fadhilah; Haerani, Elin; Vitriani, Yelvi; Yusra, Yusra
Bulletin of Computer Science Research Vol. 6 No. 1 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bulletincsr.v6i1.932

Abstract

Sentiment analysis is an important method for understanding user perceptions of cryptocurrency assets such as Bitcoin, whose price movements are strongly influenced by public opinion. This study aims to classify user sentiment from comments posted on the X platform into two classes, namely positive and negative, using the Decision Tree C4.5 algorithm. The dataset consists of 5,000 Indonesian-language comments collected through a web scraping process and processed through text preprocessing and TF-IDF–based feature extraction. The model was trained using a 70% training data and 30% testing data split. The evaluation results show that the C4.5 model achieved an accuracy of 78%. For the positive class, the model obtained a very high recall of 0.99 with an F1-score of 0.83, indicating strong performance in identifying positive comments. In contrast, the negative class achieved a recall of 0.51 with an F1-score of 0.67, despite having a high precision of 0.97. The disparity in performance between classes is influenced by the data distribution, which is not fully balanced, with positive comments being more dominant than negative ones, causing the model to be more sensitive to the majority class. Overall, the results indicate that the Decision Tree C4.5 algorithm is sufficiently effective for Indonesian-language Bitcoin sentiment classification, although it still has limitations in recognizing the minority class. Future research may explore the application of data imbalance handling techniques or more advanced algorithms to improve the balance of classification performance across classes.
Implementasi Metode RBMT dalam Penerjemahan Bahasa Indonesia ke Bahasa Makassar Hanif, Wan Muhammad; Yusra, Yusra; Muhammad Fikry; Febi Yanto; Siska Kurnia Gusti
Bulletin of Computer Science Research Vol. 6 No. 1 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bulletincsr.v6i1.935

Abstract

?This research was conducted to address the limited availability of linguistic resources for regional languages, particularly Makassar Language, which does not yet have adequate automatic translation support. The main problem addressed in this study is the absence of a reliable automatic translation system for Makassar Language. The objective of this research is to apply a rule-based translation method to translate text from Indonesian into Makassar Language. This study focuses on the implementation of the Rule-Based Machine Translation (RBMT) method for translating Indonesian text into Makassar Language using the Python programming language. The RBMT implementation involves tokenization, morphological analysis, vocabulary matching, and the application of grammatical rules, including the identification of prefixes and suffixes. The data used consist of a bilingual dictionary compiled from various sources and a set of test sentences representing everyday sentence structures. Translation evaluation was carried out using the Word Error Rate (WER) method, yielding a result of 0.289, and the Character Error Rate (CER) method, with a result of 0.21, which fall into the “Good” category based on the evaluation scale. The main findings indicate that the application of the RBMT method is capable of producing reasonably accurate translations at both the word and character levels. These findings demonstrate that a rule-based approach can be effectively applied to regional languages with limited digital data and provide an initial overview of the potential use of rule-based methods to support the development and preservation of regional languages.
Penerapan Integrasi Metode AHP dan Entropy dalam Pengambilan Keputusan Seleksi Beasiswa Perguruan Tinggi Permadi, Yuda; Saprudin, Saprudin; Rosyani, Perani
Bulletin of Computer Science Research Vol. 6 No. 1 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bulletincsr.v6i1.936

Abstract

Multi-criteria decision making (MCDM) is a complex process involving multiple criteria with different levels of importance. One of the most critical aspects of MCDM is the determination of criterion weights, as inappropriate weighting may lead to biased and unrepresentative decision outcomes. This study aims to apply an integrated weighting approach using the Analytic Hierarchy Process (AHP) and the Entropy method in multi-criteria decision making. The research employs a case study of scholarship recipient selection at University X, using decision data consisting of five student alternatives and four evaluation criteria, namely Grade Point Average (GPA), parents’ income, organizational activities, and essay assessment. The AHP method is used to derive subjective weights based on decision-makers’ preferences, while the Entropy method is applied to obtain objective weights based on data variability. Weight integration is performed using the arithmetic mean approach to produce final weights that balance subjective judgment and objective information. The results indicate that the integrated AHP–Entropy weights produce a more balanced distribution compared to single-method weighting. Weight stability is evaluated through comparative analysis of weight variations across methods, demonstrating that the integration reduces subjective dominance and enhances the representation of objective data characteristics. Therefore, the integration of AHP and Entropy is proven to be an effective weighting approach for multi-criteria decision making, particularly in the context of scholarship selection in higher education institutions.

Page 31 of 31 | Total Record : 310

 27   28   29   30   31 

Filter by Year

2020 2025


Reset
Filter By Issues
All Issue Vol. 6 No. 1 (2025): December 2025 Vol. 5 No. 6 (2025): October 2025 Vol. 5 No. 5 (2025): August 2025 Vol. 5 No. 4 (2025): June 2025 Vol. 5 No. 3 (2025): April 2025 Vol. 5 No. 2 (2025): February 2025 Vol. 5 No. 1 (2024): December 2024 Vol. 4 No. 6 (2024): Oktober 2024 Vol. 4 No. 5 (2024): August 2024 Vol. 4 No. 4 (2024): June 2024 Vol. 4 No. 3 (2024): April 2024 Vol. 4 No. 2 (2024): Februari 2024 Vol. 4 No. 1 (2023): December 2023 Vol. 3 No. 6 (2023): October 2023 Vol. 3 No. 5 (2023): Agustus 2023 Vol. 3 No. 4 (2023): June 2023 Vol. 3 No. 3 (2023): April 2023 Vol. 3 No. 2 (2023): Februari 2023 Vol. 3 No. 1 (2022): Desember 2022 Vol. 2 No. 3 (2022): Agustus 2022 Vol. 2 No. 2 (2022): April 2022 Vol. 2 No. 1 (2021): Desember 2021 Vol. 1 No. 3 (2021): Agustus 2021 Vol. 1 No. 2 (2021): April 2021 Vol. 1 No. 1 (2020): Desember 2020 More Issue