Jurnal Sistemasi (OJS FTIK - UNISI, Fakultas Teknik dan Ilmu Komputer Universitas Islam Indragiri)
Not a member yet
946 research outputs found
Sort by
Implementation of Role-Based Access Control (RBAC) in a Drug Stock Management Information System
Full text linkRole-Based Access Control (RBAC) is a role-based access management mechanism that restricts user privileges according to their authority within an information system. The implementation of this mechanism is particularly important in pharmacy drug stock management systems, especially at Apotek Pratama dr. Moris, which still relies on manual stock recording. This condition often results in data discrepancies, delays in monitoring expired medications, difficulties in report generation, and the absence of clear access restrictions for each user. This study focuses on developing a web-based drug stock management information system with RBAC as the primary mechanism for user authorization and security. The system was developed using the Waterfall methodology, which includes requirement analysis, system design, implementation, and testing. The application was built using Laravel framework version 11, MySQL database, and the Laravel Spatie Permission package for managing roles and permissions. Black Box Testing results indicate that all functional test scenarios were executed successfully, achieving a 100% success rate. User acceptance testing, conducted using the System Usability Scale (SUS), yielded an average score of 76, categorized as Good. The findings demonstrate that the implementation of RBAC effectively restricts user access based on roles, enhances data security, and improves the accuracy and efficiency of drug stock management compared to the previous manual system
House Price Prediction using the Random Forest Regression Algorithm
Full text linkHouse price prediction is a complex problem because it is influenced by various factors such as building quality, location, and living area size. As a result, conventional methods often lack accuracy in estimating housing prices. This study aims to apply the Random Forest Regression (RFR) algorithm to predict house prices using the House Prices – Advanced Regression Techniques dataset from Kaggle, which contains 1,460 property records. The SEMMA (Sample, Explore, Modify, Model, Assess) methodology was adopted due to its systematic workflow and structured focus, which improves the reliability of the developed model. In the modeling stage, RFR was implemented because it is capable of handling non-linear patterns and maintains stable performance even with a large number of features. Based on the evaluation results, the model achieved a Root Mean Squared Error (RMSE) of 28,452.75 and a coefficient of determination (R²) of 89%. This was followed by a robustness test with an RMSE of 30,665.40, indicating the stability of the model. Feature importance analysis also revealed that OverallQual had the greatest influence on house price prediction. These findings confirm that Random Forest Regression is a reliable method for predicting house prices and has strong potential to be further developed for price recommendation systems, automated property valuation, and integration into digital platforms within the real estate industry
Automation of Water Quality Recovery in Vannamei Shrimp Aquaculture using the KNN Algorithm and Fuzzy Logic
No full textVannamei shrimp play a crucial role in Indonesia’s fisheries and export industries. Despite their high potential, shrimp aquaculture still faces significant challenges, particularly disease susceptibility caused by fluctuations in water quality and pond environmental conditions. This study aims to develop a system capable of automatically monitoring and restoring water quality using the K-Nearest Neighbors (KNN) algorithm and fuzzy logic. The research adopts a research and development (R&D) approach, which includes problem analysis, data collection, system design, development, testing, evaluation, and implementation. The system employs the KNN algorithm with K=5K = 5K=5 to diagnose water quality conditions, while fuzzy logic is used to automatically control aerators, pumps, and drainage systems. The sensors utilized include salinity, pH, temperature, dissolved oxygen, and turbidity, all integrated through an ESP32 microcontroller within an Internet of Things (IoT) network. The results demonstrate that the system achieves a diagnostic accuracy of 95% and is capable of automatically controlling recovery devices. With real-time and automated operation, the system effectively maintains pond water quality, thereby improving productivity and the overall success of vannamei shrimp aquaculture
Virtual Graph Modeling for Smart Building Energy Analysis from Sequential IoT Data
Full text linkThis study proposes and validates a cost-effective sequential energy mapping methodology to address the challenges of implementing expensive energy monitoring systems in smart buildings. By utilizing a single Internet of Things (IoT) sensor deployed sequentially across five strategic locations, a comprehensive Energy Profile Repository was successfully developed. Data analysis across various scenarios accurately identified major energy hotspots, with the computer laboratory recording the highest consumption at 2.80 kWh/hour. The study also quantified potential energy waste from “vampire loads,” estimated at 2,190 kWh/year at a single location. Furthermore, the analysis revealed that the heating, ventilation, and air conditioning (HVAC) system is the dominant contributor (65%) to peak load demand, rather than computer units. The individual energy profiles were subsequently synthesized to construct a virtual energy graph that models the characteristics and structural relationships of the building’s energy network. This methodology is proven to be an effective and cost-efficient foundational approach for generating actionable, data-driven insights, while also providing a valid basis for the future development of more advanced real-time energy management systems
Comparative Analysis of CNN and Random Forest for Cashew Plant Disease Classification
Full text linkCashew plants are a strategic commodity in Indonesia and are highly susceptible to various diseases, making fast and accurate identification techniques essential to minimize economic losses. This study evaluates the comparative performance of Convolutional Neural Networks (CNN) and Random Forest in classifying cashew plant disease images. Using an experimental quantitative approach, the study utilizes a dataset of 6,549 images divided into five classes: anthracnose, gummosis, healthy, leaf miner, and red rust. The models were validated using a split sampling technique with a ratio of 80:10:10 for training, validation, and testing, and were evaluated based on F1-score as well as accuracy, precision, and recall metrics. The Random Forest model employs manual feature extraction, including color, texture (Gray-Level Co-occurrence Matrix/GLCM), and shape (Hu Moments), whereas the CNN model uses a custom Sequential architecture with automatic feature extraction. The experimental results show that CNN achieves an accuracy of 85.80%, outperforming Random Forest by approximately 9%, which attains an accuracy of 76.95%. The novelty of this study lies in the integration of high-level texture features into the Random Forest model to evaluate the performance limits of conventional machine learning compared to CNN-based automatic feature extraction. The findings indicate that CNN performs better for this dataset. However, further optimization—particularly in handling natural background variations—is still required for practical deployment
Evaluation of Information Technology Governance Maturity Level using COBIT 2019 in the Academic Information System at Nahdlatul Ulama University of Lampung
Full text linkThe rapid development of Information Technology (IT) has significantly facilitated human activities in carrying out work and various other tasks, including in the education sector. The utilization of IT plays an important role in supporting administrative processes and academic activities; therefore, proper governance is required to ensure its effective and efficient use. This study employed a descriptive quantitative approach. Data were collected through observation and the distribution of questionnaires to students. The sample size was determined using the Slovin formula with a 10% margin of error, resulting in a minimum required sample of 94 respondents. In total, this study successfully collected data from 108 student respondents. This research aimed to evaluate the maturity level of IT governance in the Academic Information System (SIAKAD) at Nahdlatul Ulama University of Lampung using the COBIT 2019 framework. The results indicated that the maturity level of SIAKAD governance was at level 4 (Predictable), meaning that the processes have been consistently implemented and well documented. The highest gap was found in the APO07 domain (Manage Human Resources), while the BAI03, DSS01, DSS02, and MEA01 domains showed stable performance in accordance with student expectations
Analysis and Implementation of Linear Regression and Decision Tree Methods to Predict Sales at Rayyan Bakery, Simpang Marbau
Full text linkThe development of information technology and data analytics has encouraged business actors to leverage historical data as a basis for decision-making. In the small and medium enterprise (SME) sector, particularly in the culinary field, the ability to predict sales is a crucial aspect of production planning and stock management to ensure operational efficiency. Rayyan Bakery Simpang Marbau, as a bakery SME, faces challenges due to fluctuating sales that have traditionally been managed based on experience rather than systematic data analysis. The main problem addressed in this study is the absence of a data-driven sales prediction method that can assist the business owner in estimating sales accurately. Therefore, a predictive approach that utilizes historical sales data is required to support managerial decision-making. This study employs linear regression and decision tree methods. The analyzed data consist of historical sales records of Rayyan Bakery Simpang Marbau over a specific period. Linear regression is used to model the linear relationship between sales variables, while the decision tree captures non-linear patterns and produces easily interpretable decision rules. The performance of both methods is analyzed and compared based on the accuracy of the predictions they generate. The results indicate that both linear regression and decision tree methods can be effectively used to predict sales; however, the decision tree provides greater flexibility in capturing fluctuating sales patterns. These findings are expected to assist Rayyan Bakery in production planning and stock management, as well as serve as a reference for applying sales prediction methods in similar SMEs
Aceh Province Tourism Destination Recommendation System using Content based Filtering Method
Full text linkTourists often experience difficulties in finding tourist destinations in Aceh Province that match their content preferences and are geographically close to their location. This study aims to develop a tourism destination recommendation system in Aceh Province using a Content-Based Filtering approach with the Cosine Similarity algorithm and the Haversine Formula. The dataset consists of 119 tourist destinations, including attributes such as destination name, destination description, and geographical coordinates (latitude and longitude). The research process began with text data preprocessing, which included case folding, punctuation removal, tokenization, duplicate word removal, stopword removal, and stemming. Next, the similarity between destinations was calculated using the Cosine Similarity algorithm based on tourism content descriptions, while the Haversine Formula was applied to measure the geographical distance between the user’s location and the tourist destinations. The results indicate that the developed system is able to provide relevant tourism destination recommendations by simultaneously considering content relevance and geographical proximity. Therefore, the system can assist tourists in selecting destinations that best match their preferences
Design of the 'Abdimas' Marketplace System: A Digital Platform for PKM Collaboration (Version 0.0)
Full text linkCommunity Service (Pengabdian kepada Masyarakat or PkM) is one of the three core responsibilities (Tridharma) of higher education in Indonesia, alongside education and research. However, many lecturers especially those new to academia face difficulties in identifying suitable community partners. This study addresses that issue through the design of a digital platform called “Abdimas”, intended to function as a marketplace system for matching lecturers with PkM partners. Applying the Design Thinking methodology, focusing on the conceptual and architectural design of the “Abdimas” platform, this study implemented a user-centered design approach. Data from interviews with lecturers, partners, and university administrators were analyzed using thematic analysis to identify core requirements. Based on these findings, the system was designed to support lecturer registration, partner discovery, and need-based matching. Senior lecturers are also supported in exploring new, underserved areas, while partners can publicly express their needs to attract suitable academic collaborators. University administrators can monitor the distribution of PkM activities over time to ensure equity and effectiveness. Unlike existing administrative platforms that often function as one-way reporting tools, the “Abdimas” marketplace introduces a bidirectional matching mechanism that allows partners to actively broadcast their specific community needs, bridging the information gap for lecturers. The system design includes use case diagrams, UI/UX prototypes, an Entity Relationship Diagram (ERD), and blackbox test scenarios to validate functionality. Although still in the design phase (Version 0.0), “Abdimas” has the potential to scale beyond academic users by supporting Corporate Social Responsibility (CSR) initiatives and facilitating student-level community services. This research contributes a structured and scalable system design to improve collaboration and outreach in community service programs within higher education
Prediction of Unpaid Student Fees at Muhammadiyah Ahmad Dahlan University Cirebon using the Random Forest Algorithm
Full text linkThis study aims to develop a predictive model for student fee payment arrears at Universitas Muhammadiyah Ahmad Dahlan Cirebon using the Random Forest algorithm. The dataset was obtained from the Academic Information System and consisted of 490 student records from four cohorts (2018–2021), which were divided into 80% training data and 20% testing data. The data processing stages included data cleaning, transformation, and feature selection using Recursive Feature Elimination (RFE). The model was optimized using GridSearchCV to obtain the best configuration. The evaluation results indicate strong performance, with an AUC of 0.980, accuracy of 88.8%, precision of 90.4%, recall of 88.8%, and an F1-score of 0.875. Feature importance analysis identified the amount of arrears variable as the most dominant factor influencing prediction outcomes. Strategic recommendations for university implementation include: (1) deploying a data-driven early warning system to identify at-risk students, (2) offering payment relief or installment programs for students with high arrears, and (3) conducting regular financial monitoring through a dashboard to support timely decision-making. Therefore, this study not only produces an effective predictive model but also provides practical solutions for improving university financial management