cover
Contact Name
Mesran
Contact Email
mesran.skom.mkom@gmail.com
Phone
-
Journal Mail Official
jurnal.bits@gmail.com
Editorial Address
-
Location
Kota medan,
Sumatera utara
INDONESIA
Building of Informatics, Technology and Science
ISSN : 26848910     EISSN : 26853310     DOI : -
Core Subject : Science,
Building of Informatics, Technology and Science (BITS) is an open access media in publishing scientific articles that contain the results of research in information technology and computers. Paper that enters this journal will be checked for plagiarism and peer-rewiew first to maintain its quality. This journal is managed by Forum Kerjasama Pendidikan Tinggi (FKPT) published 2 times a year in Juni and Desember. The existence of this journal is expected to develop research and make a real contribution in improving research resources in the field of information technology and computers.
Arjuna Subject : -
Articles 889 Documents
Prediksi Penyakit Kanker Payudara Menggunakan Algoritma Synthetic Minority Oversampling Technique dan Categorical Boosting Classifier Mandala, Muhamad Bintang; Witanti, Wina; Komarudin, Agus
Building of Informatics, Technology and Science (BITS) Vol 7 No 1 (2025): June (2025)
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i1.7403

Abstract

Breast cancer remains one of the leading causes of mortality worldwide, with high prevalence rates among women in Indonesia. Accurate and efficient diagnostic models are essential to support early detection and reduce mortality. This study aims to develop a predictive model for breast cancer classification using the CatBoost algorithm, a gradient boosting method known for its ability to natively handle categorical features and reduce overfitting through ordered boosting. The dataset used consists of diagnostic features of breast tumors, which were preprocessed by checking completeness and transforming numerical attributes into categorical bins to capture value distribution more effectively. To address class imbalance between benign and malignant cases, the SMOTE (Synthetic Minority Over-sampling Technique) method was applied, resulting in a balanced training set. Optimal hyperparameters for the CatBoost model were obtained using Bayesian optimization, with key parameters including depth, learning rate, and L2 regularization. The model was then trained and evaluated using recall, accuracy, and F1-score metrics, with a confusion matrix used to assess prediction quality. The results demonstrate that CatBoost achieved high performance with a recall of 1,0, accuracy of 98,6%, and F1-score of 0,99, outperforming or matching other benchmark models such as SVM, Neural Network, and XGBoost. These findings highlight the reliability and effectiveness of CatBoost in supporting medical decision-making for breast cancer diagnosis.
Pengelompokan Tingkat Stres Akademik Pada Mahasiswa Menggunakan Algoritma K-Medoids Nurfadilah, Nova Siska; Budianita, Elvia; Nazir, Alwis; Insani, Fitri; Susanti, Reni
Building of Informatics, Technology and Science (BITS) Vol 7 No 1 (2025): June (2025)
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i1.7409

Abstract

Academic stress is one of the common problems issues by university students due to heavy with heavy workloads, grade pressure, and various academic This condition can have a negatively impact on mental health, productivity and overall academic performance. In the long term, unmaged stress may lead serious psychological disorders. Therefore, it is important to accurately identify and classify the levels of academic stress. This study aims to cluster students’ academic stress levels by utilizing the K-Medoids algorithm. The data analyzed in the research were collected through questionnaires that were filled out by 507 students from the 2021-2023 cohorts, based on a modified version of the Perception of Academic Stress Scale (PASS). The results show that the K-medoids algorithm successfully clustered the data in 2 groups: cluster 0, which represents a moderate stress level with 212 students, and cluster 1, which indicates a high stress level with 295 students. This high-stress cluster exhibited higher average cores on questions 12 and 13 (score 3-5), which fall under the favorable category and are suspected to be the main triggers of academic stress among students in this group. Based on two evalutation metrics-Silhouette Coeficient and Davies-Bouldin Index (DBI)-it can be concluded that the optimal number of clusters for this data set is K=2. However, the clustering separation was not optimal due to he variation in study programs and the uneven distribution of respondets across academic years. This research is expected to provide direction the development intervation policies and strategies to support student welfare.
Penerapan Algoritma K-Means Untuk Mengelompokkan Tingkat Stres Akademik Pada Mahasiswa Wiranti, Lusi Diah; Budianita, Elvia; Nazir, Alwis; Insani, Fitri; Susanti, Reni
Building of Informatics, Technology and Science (BITS) Vol 7 No 1 (2025): June (2025)
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i1.7410

Abstract

Academic stress is a prevalent concern among university students, often arising from various challenges within the academic environment. These challenges may include tight assignment deadlines, elevated expectations from both lecturers and parents, ineffective time management, and negative self-assessment. If left unaddressed, such stress can negatively impact students’ academic performance and mental well-being. This study focuses on categorizing student academic stress levels using the K-Means clustering algorithm. Data were collected from 507 participants through a customized version of the Perception of Academic Stress Scale (PASS) questionnaire, adapted to suit the study context. Prior to analysis, the data were preprocessed and converted into a numerical format. Clustering was performed using Python on the Google Colab platform. To assess the clustering performance, two evaluation metrics were used: the Davies-Bouldin Index (DBI) and the Silhouette Coefficient. Lower DBI values suggest that the clusters formed are more compact and distinct from each other, while higher Silhouette values indicate better clustering performance. From the evaluation, the best clustering result was found when the number of clusters was 2, with a DBI score of 1.43 and a Silhouette score of 0.27. Nonetheless, these values still fall short of the ideal range, likely due to the heterogeneous nature of the data, as participants came from five different departments within the Faculty of Science and Technology. Moreover, the number of responses varied across academic years (2021–2023). Cluster 1 comprised 229 students identified as having low levels of academic stress, as shown by their lower questionnaire scores. In contrast, Cluster 2 consisted of 278 students with higher levels of stress, as reflected in their higher scores (ranging from 3 to 5) on positively worded items.
Analisis Sentimen Rencana Penerapan Cukai Pada Minuman Manis Kemasan Menggunakan Algoritma Naive Bayes dan Logistic Regression Gozali, Gozali; Baihaqi, Kiki Ahmad; Sukmawati, Cici Emilia; Wahiddin, Deden
Building of Informatics, Technology and Science (BITS) Vol 7 No 1 (2025): June (2025)
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i1.7411

Abstract

The plan to impose excise tax on packaged sweetened beverages (PSB) is proposed as a strategic measure to reduce sugar consumption among the public. This policy has elicited various responses from society, especially on social media platforms such as TikTok. The purpose of this study is to evaluate public sentiment towards the PSB excise tax policy by analyzing comments posted on the TikTok platform, comparing the performance of the Naive Bayes and Logistic Regression algorithms. Data were collected from comments on news videos about the implementation of the excise tax on PSB posted by official journalist accounts on TikTok, using the TikTok Comments Scraper available on the apipy website, resulting in 1,332 comments. The data were processed through preprocessing steps including text cleaning, tokenization, stemming, and word weighting using TF-IDF. After expert sentiment labeling, the data were then split into training and testing sets with an 80:20 ratio. Evaluation was conducted using a confusion matrix to obtain performance metrics such as accuracy, precision, recall, and F1-score for each model. The analysis revealed that negative comments dominated at 65.2%, while positive comments accounted for 34.8%. The Logistic Regression algorithm achieved an accuracy of 81.37%, precision of 86.22%, recall of 75.14%, and an F1-score of 77.06%. Meanwhile, the Naive Bayes algorithm obtained an accuracy of 79.85%, precision of 82.19%, recall of 74.17%, and an F1-score of 75.76%. It can be concluded that the majority of TikTok users still express negative responses to the PSB excise tax policy, and the Logistic Regression algorithm demonstrates superior performance in sentiment classification compared to the Naive Bayes algorithm.
Deteksi URL Phishing Menggunakan Natural Language Processing Dan Support Vector Machine Berbasis Machine Learning Nabila, Nabila; Hesti, Emilia; Aryanti, Aryanti
Building of Informatics, Technology and Science (BITS) Vol 7 No 1 (2025): June (2025)
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i1.7443

Abstract

Phishing represents a significant danger in cybersecurity, using malicious URLs to mislead users into revealing critical information. This research seeks to create a phishing URL detection model using machine learning via the integration of structural URL feature extraction, Natural Language Processing (NLP) methodologies, and the Support Vector Machine (SVM) classification algorithm. Indicators of phishing trends are derived from features such as URL length, the quantity of dots, and slashes, while URL content is quantified as numerical vectors using Term Frequency-Inverse Document Frequency (TF-IDF). All characteristics are subsequently integrated as input into a support vector machine model with a linear kernel for classification. The evaluation results from the classification report indicate that the integration of TF-IDF and linear kernel SVM achieves optimal performance, with 90% accuracy, 92% precision, 89% recall, and 90% F1-score. Conversely, the confusion matrix reveals 90.29% accuracy, 91.66% precision, 88.62% recall, and 90.12% F1-score. This study primarily contributes by integrating NLP and SVM into a unified adaptive phishing detection model via the amalgamation of structural and textual aspects of URLs. This strategy facilitates enhanced phishing detection relative to techniques reliant only on manual characteristics. This model, unlike other research that concentrated on particular instances or excluded NLP, is engineered to identify many categories of phishing URLs broadly, hence enhancing its relevance in tackling the dynamic nature of assaults.
Perbandingan Kinerja Metode Naïve Bayes dan Random Forest untuk Klasifikasi Penyakit Diabetes Berdasarkan Data Medis Pradana, Rendy Risqi; Astuti, Yani Parti
Building of Informatics, Technology and Science (BITS) Vol 7 No 1 (2025): June (2025)
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i1.7446

Abstract

Diabetes mellitus merupakan penyakit tidak menular yang prevalensinya terus meningkat di Indonesia. Proses diagnosis secara konvensional sering menghadapi berbagai tantangan, seperti keterlambatan dan biaya yang tinggi. Penelitian ini bertujuan untuk membandingkan kinerja algoritma Naive Bayes dan Random Forest dalam klasifikasi diabetes dengan menggunakan dataset Pima Indians Diabetes. Untuk mengatasi ketidakseimbangan kelas, dataset diproses menggunakan teknik Synthetic Minority Over-sampling Technique (SMOTE). Evaluasi kinerja dilakukan menggunakan metrik akurasi, presisi, recall, dan F1-score. Hasil penelitian menunjukkan bahwa algoritma Random Forest memperoleh akurasi sebesar 79,5%, presisi 79,6%, recall 79,5%, dan F1-score 79,5%. Sementara itu, algoritma Naive Bayes memperoleh akurasi 76,5%, presisi 76,5%, recall 76,5%, dan F1-score 76,5%. Temuan ini menunjukkan bahwa Random Forest unggul dalam menangani data yang kompleks dengan akurasi prediksi yang lebih tinggi, sedangkan Naive Bayes tetap efektif untuk implementasi yang lebih sederhana karena efisiensi komputasinya. Studi ini berkontribusi dalam pengembangan sistem pendukung keputusan cerdas untuk deteksi dini diabetes yang lebih cepat dan akurat, sehingga dapat membantu mengurangi beban pada sistem layanan kesehatan.
Deteksi Penyakit Paru-Paru Berdasarkan Gambar Citra X-Ray Menggunakan Arsitektur Convolutional Neural Network (Arsitektur Mobilenetv2) Syaifurrahman, Rizky; Silmina, Esi Putri
Building of Informatics, Technology and Science (BITS) Vol 7 No 1 (2025): June (2025)
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i1.7457

Abstract

The lungs are vital organs in the respiratory system that exchange gases, such as oxygen and carbon dioxide. However, poor air quality can lead to health problems, including lung diseases such as pneumonia, pneumothorax, lung cancer, and tuberculosis. The objective of this study is to develop an automatic detection model that uses the Convolutional Neural Network (CNN) architecture, specifically MobileNetV2, to classify X-ray images into five categories: four types of lung disease and normal lungs. The dataset consists of 2,500 images, which are divided into five classes: 80% for training, 10% for validation, and 10% for testing. Preprocessing includes resizing images to 224 x 224 pixels, normalizing pixel values, and using augmentation techniques to increase data variation. The resulting model demonstrated good performance, achieving a training accuracy of 98.76% and a validation accuracy of 97.20%. Evaluation using a confusion matrix yielded an overall F1 score of 0.94, with the highest value of 0.98 for pneumothorax. These results suggest that the model can accurately detect and classify lung diseases with an overall accuracy of 94.4%. This research significantly contributes to developing an automated lung disease detection system that can be implemented in web- or mobile-based applications and performs well across all classes.
Komparasi Algoritma Naïve Bayes, Support Vector Machine, dan Random Forest Untuk Analisis Sentimen Ulasan Pengguna Aplikasi CGV Cinemas Indonesia Febriyanti, Natasya; Rozi, Anief Fauzan
Building of Informatics, Technology and Science (BITS) Vol 7 No 1 (2025): June (2025)
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i1.7459

Abstract

CGV Cinemas Indonesia is the official platform of the CGV cinema network, designed to facilitate users in accessing cinema services digitally. Google Play Store provides a review feature based on user experience, which can influence potential users in considering whether to use the application. This study aims to analyze user sentiment toward the CGV Cinemas Indonesia application using the Naïve Bayes, SVM, and Random Forest algorithms to classify sentiments as positive, negative, or neutral. In addition, this research seeks to compare the effectiveness of the three algorithms and identify which aspects of the service are most criticized and appreciated by customers. The dataset was collected through scraping Google Play using the Python programming language, resulting in 6,629 review data points. The results show that the accuracy of Naïve Bayes is 75.2%, SVM is 88.1%, and Random Forest is 85.8%, indicating that SVM is the most effective method for sentiment analysis in this study. This research is expected to help potential users understand sentiments toward the application and provide valuable insights for CGV Cinemas Indonesia to improve service quality.
Comparison of Certainty Factor and Dempster-Shafer Methods in ENT Disease Diagnosis Expert System Sianturi, Susy Katarina; Sutopo, Teguh; Satriani, Dina; Gustina, Gustina; Faozin, Ali
Building of Informatics, Technology and Science (BITS) Vol 7 No 1 (2025): June (2025)
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i1.7465

Abstract

Diagnosis of Ear, Nose, and Throat (ENT) diseases often faces obstacles in determining the level of certainty of a disease based on the symptoms experienced by the patient. The main problem in this research is how to compare the level of accuracy between the Certainty Factor and Dempster-Shafer methods in an expert system for diagnosing ENT diseases. As a solution, this research applies both methods and analyzes the results of their calculations on various symptoms entered by the patient. The purpose of this research is to determine which method is more effective in providing certainty of diagnosis. The results show that the Certainty Factor method produces a higher level of certainty than Dempster-Shafer, for example in Tonsillitis disease which reaches 94.68% compared to 0.02% in Dempster-Shafer. Thus, the Certainty Factor method is more recommended for ENT disease diagnosis expert systems. The contribution of this research is to provide insight into the use of artificial intelligence methods in the medical field, especially in improving the accuracy of expert systems to assist health workers in making diagnostic decisions.
Sentiment Analysis of Wondr by BNI App Reviews on Play Store using the CNN-LSTM Method Putra, Ihsanudin Pradana; Sibaroni, Yuliant; Prasetiyowati, Sri Suryani
Building of Informatics, Technology and Science (BITS) Vol 7 No 1 (2025): June (2025)
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i1.7477

Abstract

As the use of digital applications in banking services increases, user opinions about these applications become an important source of data to study Wondr by BNI, which receives many user reviews, is one of the applications studied in this research. This research aims to build an accurate sentiment classification model and compare the effectiveness of two word representation methods, Word2Vec and FastText, to automatically classify sentiment into two classes, positive and negative, from unstructured review text using informal language. The data was processed through pre-processing, labeling, and processing stages using a hybrid CNN-LSTM model with 20,000 reviews available on the Google Play Store. The training process is carried out using 5-fold cross-validation and hyperparameter optimization using the random search method. The results show that the model with FastText has an accuracy of 86.38%, precision of 86.82%, recall of 86.46%, and F1-score of 86.46%. In contrast, the model with Word2Vec has an accuracy of 85.90%, precision of 86.38%, recall of 85.80%, and F1-score of 85.87%. These results show that FastText is better in accuracy and performance consistency compared to Word2Vec. This research provides a better understanding of how word representation methods affect sentiment analysis in app reviews and is expected to be a reference for future development of similar models.