Journal of Computer Networks, Architecture and High Performance Computing
Not a member yet
473 research outputs found
Sort by
Analysis of Batrsiyia Product Sales Prediction Using Linear Regression Method
Full text linkThe rapidly growing herbal and health industry encourages the need for accurate sales planning to avoid the risk of shortages or excess stock. This research aims to predict sales of Batrsiyia products using the Linear Regression algorithm with RapidMiner tools, through analyzing historical data such as sales time, number of products sold, and unit prices to identify patterns and trends to produce accurate predictions. The results show that the Linear Regression algorithm is able to predict sales with an RMSE value of 96687030.354 +/- 0.000, and a Squared Error of 9348381838748252.000 +/- 25081062946532056.000. This approach helps companies understand sales patterns, predict future trends, and optimize stock and marketing strategies. By utilizing data mining-based prediction methods, companies can make more informed decisions in meeting customer needs, maintaining business stability, and improving operational efficiency
Sentiment Analysis on Teacher Salary Policy in Indonesia 2025 Using Support Vector Machine: A Case Study on Instagram Data
Full text linkThe government's policy regarding salary increases for both civil servant (ASN) and non-civil servant (honorary) teachers in 2025 has generated various responses from the public, especially on social media. This issue has sparked public debate, with the emergence of both positive and negative comments, particularly on the Instagram platform. This study employs the Support Vector Machine (SVM) approach to classify public sentiment based on Instagram comments. A total of 1,500 comments were collected from the @folkative account during December 2024. The data were analyzed through several preprocessing stages (cleaning, case folding, tokenization, filtering, stopword removal, and stemming), followed by TF-IDF word weighting, normalization, and SVM model training and testing with an 80% training and 20% testing data split. The developed model demonstrated excellent performance, achieving an accuracy of 86%, precision of 87%, recall of 99%, and F1-score of 93%. These results indicate that the SVM algorithm is effective in classifying public opinion on government policies. This research also contributes to the advancement of machine learning applications in policy analysis based on public opinion, which can serve as valuable input for formulating more responsive policies
Comparison of Support Vector Machine and Naïve Bayes to Sentiment Analysis of Military Barracks Program
Full text linkSentiment analysis is a study that analyzes a person's emotions about a problem. The military barracks program proposed by the Governor of West Java has drawn pros and cons from the community, especially in application X. Some people consider this program to be the right solution to discipline and shape the character of students, others think that the program can take away children's freedom and rights and does not guarantee any change in character after the students leave the military barracks. Therefore, a sentiment analysis was conducted with the aim of understanding public sentiment and comparing the accuracy of the SVM and Naïve Bayes in predicting public sentiment towards the military barracks program. The method in this study begins with data crawling, data selection, labeling, data preprocessing (data cleaning, normalization, case folding, stopword removal, tokenizing, stem), TF-IDF, Word Cloud, classification with Naïve Bayes and SVM, ending with a Confusion Matrix. In contrast to SVM, which revealed that 1429 tweets had positive sentiment and 447 had negative sentiment, Naïve Bayes results indicated that 1309 tweets had positive sentiment and 567 had negative sentiment. The accuracy value of the Naïve Bayes was 91.24%, the precision was 99.73%, and the recall was 82.94%. In contrast, the SVM achieved 92.16%, the precision was 97.86%, and the recall was 86.40%. Based on these findings, it can be said that the SVM is more accurate than Naïve Bayes and that the public generally has a favorable opinion of the military barracks program
Prediction of Ovarian Cyst Disease Mortality Rate Cases Using Markov Chain Monte Carlo with Gibbs Sampling Algorithm
Full text linkOvarian cyst is one of the reproductive disorders that can develop into ovarian cancer and cause death if not treated properly. This study aims to predict the death rate due to ovarian cyst disease using the Markov Chain Monte Carlo (MCMC) method with the Gibbs Sampling algorithm. The data used is secondary data from Malahayati Islamic Hospital Medan City in 2024, which consists of 15 patients, including one deceased patient (fictitious) for the purposes of the classification model. The independent variables used include age, length of hospitalization, and number of diagnoses, while the dependent variable is the patient's death status. The estimation process was conducted with 600 iterations, where the initial 100 iterations were used as burn-in, and the rest were used to obtain the posterior mean of the model parameters. The results show that the model is able to predict death status with 100% accuracy, where all predictions match the actual data. The parameter coefficients show that the higher the age, the longer the hospitalization, and the more the number of diagnoses, the higher the risk of death. The MCMC method with Gibbs Sampling algorithm proved to be effective in generating probabilistic predictions as well as identifying important factors that affect the risk of death of patients with ovarian cyst
STRATEGIC INFORMATION SYSTEMS PLANNING USING THE WARD AND PEPPARD METHOD (A CASE STUDY OF KOPERASI DAUH AYU)
Full text link Koperasi Dauh Ayu requires a formally articulated IS/IT strategy to overcome fragmented, manual operations and move toward an integrated, member-centric model. This case study applies the Ward and Peppard framework to diagnose business and IS/IT conditions, using PEST, Porter’s Five Forces, value chain, technology trend scanning, and SWOT with quantitative IFAS–EFAS scoring from six expert respondents. The cooperative is positioned in Quadrant I of the SWOT map with coordinates X = 0.309 and Y = 0.397, indicating an aggressive strategy space where internal strengths can be leveraged to seize external opportunities. The study produces a prioritized portfolio of fourteen applications mapped with the McFarlan Grid, alongside an IT strategy for network, hardware, and platform modernization, and an IS/IT management strategy that establishes a dedicated ICT unit and governance mechanisms. Recommended initiatives are expected to reduce cycle times and error incidence, consolidate a single source of truth for member and financial data, and elevate service quality. The contribution extends the application of Ward and Peppard to the cooperative sector in Indonesia, a context less examined than large enterprises, and shows how staged capability building can translate environmental enablers into realized digital benefits. Limitations include a single-case design without post-implementation measurement; future work should pilot priority systems and evaluate pre–post performance and cost–benefit outcomes
Development of the ENC Android Application for Electronic Nursing Care
Full text linkThe rapid digitalization of the healthcare sector is transforming nursing education and practice, particularly in the area of clinical documentation. In Indonesia, as mandated by Peraturan Menteri Kesehatan Nomor 24 Tahun 2022, all healthcare facilities must document patient medical histories electronically. However, nursing students face challenges in practicing electronic documentation due to limited access to hospital-based EMR systems. This study aims to develop an Electronic Nursing Care (ENC) application as an Android-based platform for nursing students to practice digital documentation aligned with national nursing standards, such as SDKI (Indonesian Nursing Diagnosis Standards), SLKI (Indonesian Nursing Outcomes Standards), and SIKI (Indonesian Nursing Interventions Standards). The ENC application was developed using an Agile methodology, consisting of five sprints: backend development, core features, UI/UX integration, advanced features, and comprehensive internal testing. A 2023 prototype trial demonstrated high user satisfaction, with 93% of students expressing satisfaction with the app’s user interface and experience. The ENC application was designed to reduce the administrative burden on nursing students by offering tools for documenting the nursing process, including assessment, diagnosis, planning, implementation, and evaluation. The initial results indicate that the application meets technical specifications and has the potential to enhance nursing education by providing a structured, user-friendly platform for digital documentation. The study suggests further pilot testing in clinical settings, development of additional features, and ongoing evaluation to ensure the application’s success in nursing education and clinical practice
SI-Konseling For Analyzing The Effect Of Stress Levels On Students' Academic Using K-Means Algorithm
No full textAlong with the rapid development of information technology, the world of education is also developing towards the digitalization era. Therefore, the use of information technology is needed to support the progress of the world of education. At SMA Negeri 6 Pematangsiantar, students who receive counseling guidance are students who do not obey school regulations such as undisciplined behavior in learning activities and extracurricular activities. According to psychologists, there is a relationship between negative behavior and the level of stress experienced by children. Signs that a child is experiencing stress include anger, aggressive behavior, and disobedience. It could be concluded that one of the triggers for negative behavior in children at school can be caused by stress. This study was conducted to determine the effect of student stress on their academic grades. By applying Data Mining using K-Means Clustering Algorithm, students can be grouped based on stress levels and academic achievement. So that a potential relationship can be found between these variables. In addition, the use of a counseling information system can improve the implementation of counseling to be more effective and efficient. Based on the research results, by collecting 960 student samples that will be used as calculation samples, the value of cluster one (C1) is at the average academic value (sufficient) with a moderate stress level percentage and cluster two (C2) at the average academic value (good) with a low stress level percentage. So it can be concluded that the percentage of stress experienced can affect their academic value
Optimization of the Shortest Route Using the Djikstra Algorithm to the Nearest Covid-19 Referral Hospital for Communities Exposed to the District of Medan Baru
Full text linkAbstract: Finding the shortest route is a problem to find a path that connects two nodes with the least amount of weight. Many methods are used in finding the shortest route. One of the methods used is Dijkstra's algorithm. Dijkstra's algorithm is an excellent algorithm used in determining the shortest route from a startingpoint toan end point (destination). In this study, the determination of the shortest route from each kelurahan in the Medan Baru District to the nearest Covid-19 referral hospital can be searched maximally using the Dijkstra algorithm with the distance taken through the google maps application. However, there are some limitations that are limitations in this study. The drawbacks are traffic jams, traffic lights, one-way streets. This cannot be ignored on routes in urban areas. In the future, researchers will look for optimization of determining the shortest route by including some of the problem constraints that occur. The Dijsktra algorithm is an application that must be modernized for more complex constraints
FUZZY LOGIC-CONTROLLED IOT SYSTEM FOR SMART PUBLIC TOILETS: DESIGN, IMPLEMENTATION, AND EVALUATION
Full text linkEfficient energy management in public facilities such as public toilets has become an important challenge in the modern era, especially with the increasing demand for environmental sustainability. In this research, we developed a smart toilet system based on the Internet of Things (IoT) using the ESP32 as the main microcontroller and fuzzy logic methods for intelligent decision-making. This system is equipped with temperature (DHT22), humidity, and distance (HC-SR04) sensors to detect environmental conditions and user presence. Based on this data, the toilet fan and light are automatically controlled to minimize energy consumption. To facilitate real-time monitoring and threshold control, this system is integrated with a Flutter-based application, which provides an intuitive user interface for viewing environmental data and setting temperature, humidity, and distance thresholds. Fuzzy logic is used to determine the fan speed based on temperature and humidity inputs, with the output being a gradual fan speed control. (PWM). The test results show that the system can reduce energy consumption by up to 30% compared to the manual method, especially by reducing the unnecessary device idle time. Additionally, the system has an average response time of 200 ms to send sensor data to the application and receive threshold updates from the user. With this approach, the research shows that the integration of IoT with fuzzy logic provides significant energy efficiency and enhances the user experience. This research also opens up opportunities for further development, such as the integration of machine learning technology for predicting facility usage patterns or the implementation of additional sensors for air quality detection. These findings support the implementation of IoT-based automated systems in public facilities to achieve energy efficiency and environmental sustainability
RAD-Based Public Opinion Monitoring Information System for BSN
Full text linkThe growing influence of online media in shaping public opinion has driven government institutions to modernize their monitoring and communication systems. This study aims to develop a web-based information system for monitoring public opinion, tailored to the needs of the National Standardization Agency of Indonesia (BSN). Using the Rapid Application Development (RAD) approach, the system was built through a phased prototyping and user involvement to ensure functional relevance. The final system enables sentiment classification of news articles, centralized data storage, trend visualization, and automated news clipping. Evaluation results indicate improvements in monitoring speed, accuracy, and usability compared to previous manual methods. This study confirms the effectiveness of RAD in building practical digital tools for public sector communication and reputation management