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 926 Documents
Analisis Sentimen Pengguna pada Aplikasi Tokopedia Menggunakan Algoritma Convolutional Neural Network Maskhuri, Alip; Hardiani, Tikaridha
Building of Informatics, Technology and Science (BITS) Vol 6 No 4 (2025): March 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i4.6923

Abstract

The Covid-19 pandemic in 2020 accelerated digital transformation across various sectors, including e-commerce. Tokopedia, Indonesia's largest e-commerce platforms, has experienced significant dynamics in user reviews that can impact its reputation. This study aims to analyze the sentiment of Tokopedia user reviews collected from the Google Play Store and the social media platform X using the Convolutional Neural Network (CNN) algorithm. The research is motivated by the increasing competition in the e-commerce industry, requiring companies to understand consumer sentiment to improve their services. The methodology includes data collection through text mining, data preprocessing, automatic labeling using the Pre-Trained IndoBERT model, and splitting the dataset into training, validation, and testing sets. A total of 15,751 reviews were sentiment with 8,885 classified as negative, 3,860 as neutral, and 3,006 as positive. The CNN algorithm was applied to classify these reviews, and the results showed that the model achieved an accuracy of 83%. The model performed best in recognizing negative sentiment but struggled to distinguish between neutral and positive sentiments due to data imbalance. This study recommends collecting more data to achieve a balanced class distribution and exploring pre-trained models such as IndoBERT or IndoNLU to enhance sentiment analysis accuracy.
Sistem Monitoring Ketinggian Permukaan Air dan Pengaturan Otomatis Pintu Air Berbasis Internet of Things Aprianti, Nurhalisa; Paniran, Paniran; Suksmadana, I Made Budi
Building of Informatics, Technology and Science (BITS) Vol 6 No 4 (2025): March 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i4.6924

Abstract

Rainfall refers to the amount of precipitation that falls in a specific area over a certain period, typically measured in millimeters (mm). High rainfall can increase the risk of flooding. This study aims to develop an IoT-based system to detect water levels, measure rainfall, and automatically control the opening and closing of floodgates. The collected data is permanently stored in a database and can be monitored remotely. The researchers utilized an Arduino Mega 2560 microcontroller and an ESP32 to process data, an ultrasonic sensor to measure water level (in cm), and a rainfall sensor (in mm). The data is transmitted to a Firebase database and connected to a Telegram bot, allowing administrators to monitor it in real time. The system classifies water levels into four categories: safe (0–30 cm), alert (31–40 cm), warning (41–50 cm), and danger (51–60 cm). Rainfall intensity is also categorized into four levels: light (0.5–20 mm/h), moderate (20–50 mm/h), heavy (50–100 mm/h), and extreme (more than 100 mm/h). The system automatically adjusts the floodgates based on water conditions but also allows for manual control. The sensors demonstrated an average accuracy of 97%, with an average notification delay of 675 ms during extreme rainfall conditions.
Political Comperative Analysis of Indonesian Political Fake News Detection using IndoBERT-Bi-GRU-Attention Models: Evaluating Performance on Narratives and News Headlines Datasets Manurung, Juliana Damayanti; Purba, Ronsen
Building of Informatics, Technology and Science (BITS) Vol 6 No 4 (2025): March 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i4.6938

Abstract

The instant and massive spread of fake news on social media negatively impacts public trust in the media and news agencies. In politics, fake news is often used by politicians to gain support ahead of elections. Detecting fake news in Indonesia poses a significant challenge, especially for communities vulnerable to misinformation. This study aims to develop a new model that combines IndoBERT with Bi-GRU and Attention. Additionally, a comparison is made between the main model and two word embedding models, FastText and GloVe. The tests were conducted on datasets of headlines and news narratives separately. Data was sourced from CNN, Tempo.co, Kompas, and TurnBackHoax.ID. The results show that the IndoBERT-Bi-GRU-Attention model with FastText excelled on the headline dataset with an accuracy of 99.76% and an F1-Score of 99.61%, while the main IndoBERT-Bi-GRU-Attention model excelled on the narrative dataset with an accuracy of 99.08% and an F1-Score of 98.40%. This research demonstrates that IndoBERT can be combined with Bi-GRU, significantly contributing to the development of fake news detection models.
Analisis Sentiment Terhadap Diabetes Menggunakan Algoritma Naïve Bayes, Random Forest, SVM Pada Media Sosial X Apriani, Linda; Hendrastuty, Nirwana
Building of Informatics, Technology and Science (BITS) Vol 6 No 4 (2025): March 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i4.6941

Abstract

Diabetes is one of the chronic diseases that has received widespread attention in society, especially on social media X. This is due to the increasing number of sufferers every year. Based on data from the World Health Organization (WHO), in 2021 it is estimated that 537 million people aged 20-79 years are living with diabetes, an increase from the 2019 estimate of 463 million people. In addition, around 1.3 million deaths are caused by diabetes, with 4 percent of them occurring before the age of 70. This condition occurs due to high blood sugar levels that interfere with the body's metabolic functions, making it difficult for the body to process sugar optimally. This study aims to compare the performance of Naïve Bayes, Random Forest, and Support Vector Machine (SVM) algorithms in sentiment analysis related to diabetes. The research data was obtained from the Twitter platform with a total of 8,401 tweets collected using crawling techniques using certain keywords in the time span of 2024 to 2025. The data then went through a pre-processing stage to produce clean data. Tests were conducted to evaluate the accuracy of each model in predicting public sentiment. The test results show that the SVM algorithm provides the best performance with 85% accuracy, followed by Random Forest with 82% accuracy, and Naïve Bayes with 74% accuracy before the application of Synthetic Minority Oversampling Technique (SMOTE). After optimization using SMOTE, the SVM algorithm still showed the best performance with 96% accuracy, followed by Random Forest with 95% accuracy, and Naïve Bayes with 85% accuracy. Based on these results, SVM proved to be the most effective algorithm in classifying sentiment related to diabetes. It is hoped that the results of this research can contribute to efforts to manage diabetes through a better understanding of public perceptions.
Klasifikasi Sentimen Ulasan Produk pada Platform E-Commerce di Indonesia dengan Menggunakan Model Pre-Trained IndoBERT Aji, Bayu Puspito; Aditya, Christian Sri Kusuma
Building of Informatics, Technology and Science (BITS) Vol 6 No 4 (2025): March 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i4.6968

Abstract

In today's digital era, sentiment analysis of product reviews on e-commerce platforms is becoming increasingly important, especially on Tokopedia, one of the largest marketplaces in Indonesia. Tokopedia provides facilities for users to leave reviews after making transactions, which play an important role in helping businesses understand customer perceptions of products. This research aims to classify the sentiment of product reviews on Tokopedia using the IndoBERT model and evaluate its performance compared to LSTM-based methods combined with FastText, Glove, and Word2Vec embedding. The LSTM-FastText model in previous research achieved the highest accuracy of 85.08%. In this study, the sentiment classification of product reviews on Tokopedia was carried out with a total of 5400 data and the sentiment classification process was divided into two categories, namely positive and negative, with the division of the dataset into three groups: training, validation, and testing. The contribution in this research is to explore the effectiveness of the IndoBERT model performance compared to previous methods that implement the LSTM model with FastText, Glove, and Word2Vec embedding. Based on the research results, the IndoBERT model achieved an accuracy of 97%, with the same F1-score value for both sentiment categories of 97%. Specifically designed with pre-training on a large Indonesian corpus, IndoBERT is able to understand the context of the text better than the LSTM model used in previous studies. This allows IndoBERT to produce higher accuracy, as it can understand product reviews in Indonesian more effectively.
Perbandingan Metode Naive Bayes, Random Forest dan SVM Untuk Analisis Sentimen Pada Twitter Tentang Kenaikan Gaji Guru Yuniar, Eny; Hendrastuty, Nirwana
Building of Informatics, Technology and Science (BITS) Vol 6 No 4 (2025): March 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i4.6970

Abstract

The increase in teacher salaries has become a highly debated issue within the community, with various opinions being expressed through social media, particularly Twitter. This study aims to analyze public sentiment regarding the teacher salary increase policy using three machine learning algorithms: Naïve Bayes, Random Forest, and Support Vector Machine (SVM). The data used consists of 6010 tweets collected on the topic, which were processed into 5531 data points after cleaning and preprocessing. This study evaluates the performance of each algorithm using accuracy, precision, recall, and F1-score metrics. The results show that SVM achieved the highest accuracy (86%) before applying the SMOTE technique, followed by Random Forest (85%) and Naïve Bayes (84%). After applying SMOTE to address data imbalance, Random Forest showed a significant performance improvement, with accuracy reaching 99%, followed by SVM (98%) and Naïve Bayes (89%). These results indicate that the SMOTE technique can effectively improve model performance, particularly in handling the imbalance between positive, negative, and neutral sentiment data. This study provides new insights into how the public responds to the teacher salary increase policy, while also introducing the use of SMOTE to enhance accuracy in sentiment analysis on social media.
Analisis Perbandingan Naïve Bayes dan Neural Network dalam Klasifikasi Minat Masyarakat pada Kursus Komputer Fitria, Nabila Syah; Suryadi, Sudi; Nasution, Fitri Aini
Building of Informatics, Technology and Science (BITS) Vol 6 No 4 (2025): March 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i4.6999

Abstract

In the digital era, the use of technology in education is growing, especially in improving people's digital literacy through computer courses. To analyze people's interest in courses, a data mining-based approach is needed that can process large amounts of data and identify certain patterns. Naïve Bayes and Neural Network are two widely used classification methods, where Naïve Bayes works based on independent probabilities between features, while Neural Network uses artificial neural networks to capture more complex patterns. This study aims to compare the two methods in classifying people's interest in LKP Ibay Komputer and evaluate the accuracy of each model. The classification results show that both methods produce the same predictions, namely 53 data are categorized as interested and 20 data as not interested. The model accuracy reaches 100%, indicating very high classification performance. Although these results seem ideal, perfect accuracy like this often raises questions regarding the validity and robustness of the model in real-world scenarios. Factors such as relatively small dataset sizes, overly structured data patterns, or lack of variation in training data can cause results that appear too good. Therefore, it is important to conduct additional evaluations such as cross-validation or testing on different datasets to ensure that the model does not experience overfitting and remains reliable in broader predictions. With these results, it can be concluded that both Naïve Bayes and Neural Networks have optimal performance in classifying people's interest in computer courses, but the choice of method can be adjusted according to needs, where Naïve Bayes excels in computational efficiency, while Neural Networks are more adaptive to more complex data.
Optimalisasi Model BioBERT untuk Pengenalan Entitas pada Teks Medis dengan Conditional Random Fields (CRF) Nafanda, Cynthia Dwi; Salam, Abu
Building of Informatics, Technology and Science (BITS) Vol 6 No 4 (2025): March 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i4.7042

Abstract

This research evaluates the performance of various models in the Named Entity Recognition (NER) task for medical entities, focusing on imbalanced datasets. Six BioBERT model configurations were tested, incorporating optimization techniques such as Class Weight, Conditional Random Fields (CRF), and Hyperparameter Tuning. The evaluation was conducted using Precision, Recall, and F1-Score metrics, which are particularly relevant in the context of NER, especially for addressing class imbalance in the data. The dataset used is BC5CDR, which targets chemical and disease entities in unstructured medical texts from PubMed. The data was divided into three parts: a training dataset for model training, a validation dataset for model tuning, and a test dataset for performance evaluation. The dataset was split evenly to ensure unbiased model testing, leading to more accurate results that can serve as a reference for developing more efficient medical NER systems. The evaluation results indicate that BioBERT + CRF is the model with an F1-Score that reflects an optimal balance between Precision (ranked 3rd, 0.6067 for B-Chemical, 0.5594 for B-Disease, 0.4600 for I-Disease, and 0.5083 for I-Chemical) and Recall (ranked 3rd, 0.5580 for B-Chemical, 0.4491 for B-Disease, 0.5718 for I-Disease, and 0.3840 for I-Chemical) compared to other models. This model proved to be more accurate in detecting medical entities without compromising prediction precision. The model's stability is also enhanced by a smaller gap between Precision and Recall, making it the best choice for NER in medical texts. The application of early stopping techniques effectively prevented overfitting, ensuring the model learned optimally without losing generalization. With better balance in recognizing medical entities from unstructured texts, this model presents the most effective approach for NER systems in the medical domain.
Analisis Sentimen Publik terhadap Virus HMPV Berdasarkan Media Sosial X dengan Algoritma Logistic Regression Wijaya, Feri Aldi; Parjito, Parjito
Building of Informatics, Technology and Science (BITS) Vol 6 No 4 (2025): March 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i4.7053

Abstract

Human Metapneumovirus (HMPV) is a virus that affects the respiratory tract, causing flu-like symptoms such as cough, fever, and nasal congestion. This virus was first discovered in 2001 and generally causes mild infections. However, certain groups, such as children, the elderly, and individuals with weakened immune systems, are at higher risk of developing severe conditions like bronchitis or pneumonia. Based on this issue, a sentiment analysis of public responses to Human Metapneumovirus (HMPV) cases was conducted using data collected from the X platform, consisting of 10,199 tweets. The data was gathered between December 1, 2024, and January 30, 2025, using Tweet Harvest in Google Colab with the Twitter API. This study applied the Synthetic Minority Oversampling Technique (SMOTE) to address data imbalance, with an 80% to 20% split between training and testing data. The results showed that before applying SMOTE, the logistic regression algorithm had an accuracy of 83%, with precision for positive sentiment at 90%, neutral at 80%, negative at 85%, while recall for positive sentiment was 89%, neutral 89%, negative 92%. After applying SMOTE, accuracy increased to 90%, with the most significant improvement observed in positive sentiment. The precision for positive sentiment reached 90%, neutral 87%, and negative 95%, while recall for positive sentiment was 96%, neutral 90%, negative 84%. This research provides insights into the use of logistic regression algorithms in sentiment analysis related to HMPV and serves as a reference for governments and health organizations in designing more effective communication strategies and interventions.
Perbandingan Algoritma Support Vector Machine, Decision Tree, Naïve Bayes, dan Neural Network dalam Klasifikasi Email Wicaksono, Dika; Agastya, I Made Artha
Building of Informatics, Technology and Science (BITS) Vol 6 No 4 (2025): March 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i4.6949

Abstract

This study aims to compare the effectiveness of four machine learning models in email classification, namely Support Vector Machine (SVM), Decision Tree, Naive Bayes, and Neural Network. This research uses datasets obtained from the Kaggle website. The first dataset contains 18,650 phishing emails (7,328 phishing and 11,322 non-phishing). The second dataset is the result of merging two different datasets containing Indonesian spam emails, resulting in a total of 4,681 emails (2,670 spam and 2,011 non-spam). The merging was done to obtain a more representative amount of data for model evaluation. The results of the study of the two datasets above showed that the Neural Network achieved the highest accuracy with an average of 96.60%. Then, followed by SVM with an average accuracy of 96.43%. Meanwhile, Decision Tree has a fairly high accuracy with an average of 92.38%. In contrast, Naive Bayes recorded the lowest performance with an average accuracy of 90.22%. Although Neural Network has the highest accuracy, other models may be more suitable depending on the needs of the system. Models with lower accuracy, such as Naive Bayes, can be more useful in systems with computational limitations due to their efficiency. SVM offers a balance between high accuracy and computational efficiency, making it an ideal choice for systems that require optimal performance without too much computational burden. Decision Tree is superior in result interpretation, making it suitable for applications that require transparency in decision making.