MEANS (Media Informasi Analisa dan Sistem)

Analisis Sentimen Berbasis Aspek pada Ulasan Mobile JKN di Google Play Store Menggunakan SVM

2026-05-11T10:10:47+00:00

Aspect-based sentiment analysis on user reviews of the Mobile JKN application was conducted to identify user perceptions regarding interface, features and performance, as well as service aspects using the Support Vector Machine (SVM) method. The data were collected through scraping Google Play Store reviews from October 2025 to December 2025, resulting in 13,503 reviews, of which 3,659 met the criteria for aspect-based analysis. The research stages included text preprocessing, TF-IDF weighting, and classification using SVM. The evaluation was performed using two data-splitting scenarios, namely 80:20 and 70:30, to assess model performance under different proportions of training and Testing data. The results indicate that the service aspect is the most frequently discussed by users. The SVM model achieved the highest Accuracy of 96.88% for the interface aspect, 90.80% for the features and performance aspect, and 95.07% for the service aspect, with Precision and Recall values indicating good classification performance.

Perbandingan Fitur Handcrafted dan Non-Handcrafted untuk Klasifikasi Kelembapan Tanah

2026-02-25T17:05:57+00:00

Soil moisture is a crucial parameter in precision agriculture because it impacts irrigation management and crop productivity. Conventional measurement methods using physical sensors have limitations in terms of cost, maintenance, and coverage area. This study aims to compare the performance of handcrafted and non-handcrafted feature approaches in digital image-based soil moisture classification. The handcrafted approach extracts 60 color features consisting of RGB and HSV statistics and HSV histograms, then classifies them using the Random Forest algorithm. The non-handcrafted approach uses the transfer learning- based MobileNetV2 architecture to automatically learn feature representations from raw images. Both methods were tested using the same dataset and experimental protocol to ensure an objective comparison. The experimental results showed that Random Forest achieved an accuracy of 91.05%, higher than MobileNetV2's 87.65%. Confusion matrix analysis indicated that the dominant misclassification occurred between the moderate and wet classes due to the similarity in color distribution. The results show that for limited datasets and problems dominated by color characteristics, handcrafted features can provide more stable and efficient performance than deep learning models. This study emphasizes the importance of selecting a classification method based on data characteristics and computational requirements, rather than solely on model complexity.

Analisis Kinerja Jaringan Menggunakan Metode PCQ Manajemen Bandwitdth Dengan Pendekatan Network Development Life Cycle (NDLC)

2026-05-22T04:21:34+00:00

Bandwidth management is a critical aspect of computer network administration, especially in multi-user environments such as network laboratories. This study analyzes the performance of the Per Connection Queue (PCQ) method at the Network Laboratory of Dhyana Pura University, involving 36 client PCs with a 2 Mbps bandwidth limit per user. Measurements were carried out using Wireshark under peak and off-peak conditions, covering throughput, delay, packet loss, and jitter. During peak hours, throughput was 0.994 Mbps, delay was 163.196 ms, packet loss was 2.31%, and jitter was 65.288 ms. During off-peak hours, throughput was 1.354 Mbps, delay was 35.231 ms, packet loss was 0.794%, and jitter was 26.021 ms. Based on TIPHON standards, throughput and jitter are classified as Good, while packet loss and delay are classified as Very Good. These findings confirm that the PCQ method on Mikrotik distributes bandwidth fairly and efficiently to all network users.

Keyword : Management Bandwidth, NDLC, PCQ, QoS

Optimasi Pemilihan Komponen Terbaik untuk Perakitan Skateboard dengan Algoritma Genetika Berbasis Java pada Toko DSRuntul

2026-06-18T12:15:37+00:00

Along with the development of skateboarding and advances in information technology, consumers are now smarter in purchasing. Consumers prefer assembled skateboards because consumers can get skateboard components according to their wishes. The problem faced by the Dsruntul Store is that it still sells complete skateboards. This can cause consumers to be unable to choose the desired skateboard components. As for the assembly according to consumer demand, the assembly still uses the conventional method of selecting each skateboard component one by one. This problem can be overcome by the existence of an application that is able to search for the selection of skateboard component combinations from each type of skateboard component according to the consumer's budget and the desired specific component criteria effectively and efficiently. In this application, the Genetic Algorithm method will be applied, which is an optimization method that can provide alternative solutions to a problem that is adapted to the genetic process of biological organisms based on Charles Darwin's theory of evolution. The coding technique used in determining skateboard components is integer representation, the selection used is roulette wheel selection, the crossover used is one point crossover and the mutation used is random mutation.

Analisis Kinerja Tata Kelola Layanan Jaringan Berdasarkan Framework COBIT 2019 pada Universitas Muhammadiyah Bima

2026-06-24T09:06:54+00:00

Perkembangan teknologi informasi telah menjadikan layanan jaringan sebagai infrastruktur penting dalam mendukung aktivitas akademik dan administrasi di perguruan tinggi. Universitas Muhammadiyah Bima memanfaatkan layanan jaringan untuk mendukung Sistem Informasi Akademik (SIAKAD), pembelajaran daring, layanan administrasi digital, serta komunikasi internal kampus. Namun demikian, masih ditemukan beberapa permasalahan seperti ketidakstabilan koneksi internet, keterbatasan bandwidth saat jumlah pengguna meningkat, belum optimalnya monitoring jaringan, serta belum adanya evaluasi tata kelola layanan jaringan yang dilakukan secara terstruktur. Penelitian ini bertujuan untuk menganalisis kinerja tata kelola layanan jaringan berdasarkan framework COBIT 2019 serta mengidentifikasi kesenjangan antara kondisi aktual dan kondisi yang diharapkan. Penelitian menggunakan pendekatan kuantitatif deskriptif dengan metode evaluatif. Pengumpulan data dilakukan melalui observasi, wawancara, dokumentasi, dan kuesioner yang disusun berdasarkan domain COBIT 2019, yaitu DSS01 (Managed Operations), DSS02 (Managed Service Requests and Incidents), DSS04 (Managed Continuity), dan APO13 (Managed Security). Data dianalisis menggunakan metode Capability Level dan Gap Analysis. Hasil penelitian menunjukkan bahwa domain DSS01 memperoleh nilai capability level sebesar 2,8, DSS02 sebesar 2,5, DSS04 sebesar 2,3, dan APO13 sebesar 2,7. Seluruh domain berada pada Level 2 (Managed Process) dengan target Level 4 (Predictable Process). Nilai kesenjangan terbesar terdapat pada domain DSS04 sebesar 1,7, diikuti DSS02 sebesar 1,5, APO13 sebesar 1,3, dan DSS01 sebesar 1,2. Hasil penelitian menunjukkan bahwa tata kelola layanan jaringan di Universitas Muhammadiyah Bima telah berjalan dan dikelola dengan baik, namun masih memerlukan peningkatan pada aspek keberlangsungan layanan, pengelolaan insiden, keamanan informasi, dan monitoring jaringan agar mencapai tingkat kapabilitas yang diharapkan. Penelitian ini menghasilkan rekomendasi perbaikan yang dapat digunakan sebagai dasar pengembangan tata kelola layanan jaringan secara efektif, efisien, dan berkelanjutan.

ANALYSIS OF THE PERFORMANCE OF SUPPORT VECTOR MACHINE IN DETECTING NETWORK INTRUSION IN THE COMPUTER LAB OF SMKN 1 WERA

2026-06-15T14:33:56+00:00

Network management in the Village Office of Nunggi has not been optimally supported by a real-time monitoring system, resulting in limited visibility of network performance and potential disruptions in service delivery. This study implements a network monitoring system based on the Simple Network Management Protocol (SNMP) on MikroTik routers to improve the effectiveness of network utilization. The system is designed to collect and analyze network performance data such as bandwidth usage, uptime, CPU load, and traffic conditions in real time. The monitoring results are displayed through a user-friendly interface to assist administrators in detecting network problems more quickly and accurately. The implementation shows that the SNMP-based monitoring system is able to provide timely and structured network information, thereby improving network stability, efficiency, and management responsiveness. Overall, this research demonstrates that SNMP implementation on MikroTik routers is effective in supporting better network monitoring and decision-making in the Village Office environment.

PENERAPAN METODE SIMPLE ADDITIVE WEIGHTING DAN WEIGHTED PRODUCT DALAM PEMILIHAN CALON DOSEN

2026-06-05T15:20:46+00:00

The selection of qualified and competent lecturer candidates plays an important role in improving the quality of higher education. However, manual selection processes are often prone to subjectivity, making it necessary to implement a decision support system that can provide more objective and structured recommendations. This study aims to apply the Simple Additive Weighting (SAW) and Weighted Product (WP) methods in the selection of lecturer candidates at University X. The evaluation criteria include educational background, academic achievement, scientific publications, technical competence, and soft skills. The SAW method utilizes normalization and weighted summation to calculate preference values, while the WP method applies weighted multiplication to determine the final score of each alternative. The results show that both methods are capable of processing candidate data effectively and supporting objective decision-making. Based on the calculations, alternative A1 obtained the highest score and was recommended as the best lecturer candidate at University X.

Redesign UI/UX pada Modul Pembelajaran Sipandu Menggunakan Metode User Centered Design dan System Usability Scale

2026-06-21T13:45:51+00:00

Penelitian ini bertujuan merancang ulang (redesign) UI/UX modul pembelajaran Sipandu menggunakan metode User Centered Design (UCD) serta mengukur tingkat kepuasan pengguna menggunakan System Usability Scale (SUS). Masalah utama sistem lama terletak pada visualisasi dan navigasi yang kurang intuitif, sehingga menghambat pembelajaran digital. Evaluasi kuantitatif dilakukan melalui kuesioner SUS kepada 96 responden mahasiswa Universitas Dhyana Pura. Hasil pengujian menunjukkan bahwa metode UCD sukses mentransformasikan portal Sipandu menjadi lebih interaktif dan adaptif. Keberhasilan ini divalidasi oleh lonjakan skor rata-rata SUS yang signifikan, dari nilai 47,40 (unacceptable/Grade E) pada sistem lama menjadi 78,13 (acceptable/good/Grade B) pada prototipe desain baru. Integrasi pendekatan UCD dan metrik SUS terbukti efektif mengoptimalkan kualitas usability serta meningkatkan akseptabilitas sistem pembelajaran digital.

Perancangan Prototype Sistem Inventori Digital Menggunakan Barcode Pada CV. XYZ Sebagai UMKM Aksesoris Berbahan Dasar Kulit

2026-05-08T10:10:25+00:00

The impact of the inventory system at CV. XYZ, which is still managed manually, results in stock information that cannot be monitored directly by the production team. The processes of recording, checking, and reporting stock require a considerable amount of time, making it difficult to support fast decision-making. Errors in entering stock quantities may lead to stock shortages or overstocking, causing financial losses for CV. XYZ when it is unable to fulfill incoming orders. In addition, excess raw materials may result in high inventory costs. This study aims to develop a digitalized inventory system at CV. XYZ to improve the efficiency and accuracy of inventory management while minimizing raw material shortages. The method used in this study is a Warehouse Management System (WMS) by implementing digital inventory management based on barcode technology. The results show that the accuracy of inventory recording increased from 82% in the manual system to 93% after the implementation of the barcode-based system. The stock auditing process became easier, and the system was able to generate reports automatically and accurately based on scanning results

Implementasi Multimedia Augmented Reality Pada Pembelajaran Sistem Primary Control Surface Pesawat Cessna 150M

2026-07-08T01:46:21+00:00

Many learning media are used in education to teach students. These media typically include 2D and 3D images, music, video, animation, virtual reality, and augmented reality to engage and enhance student learning. This lesson explains the components of a Cessna 150M aircraft using Augmented Reality (AR) with student users. It is hoped that with this learning media, students will learn and understand the function of an aircraft's primary control surfaces. The method used is the Multimedia Development Life Cycle, which has six stages: concept, design, material collection, assembly, testing, and distribution, using AR (Marker-Based Augmented Reality). Based on testing results from the research, this learning media can be used on Android 10 and 11 smartphones and can provide information about aircraft components and is suitable for use

Penerapan Sistem Pendukung Keputusan Berbasis Kecerdasan Buatan Untuk Penentuan Prioritas Penerima Bantuan Pendidikan Menggunakan Pendekatan Hybrid AHP dan Decision Tree

2026-07-20T18:26:10+00:00

Educational assistance distribution is one of the strategic efforts made by governments and educational institutions to improve equal access to education. However, in practice, the process of determining priority recipients often faces several problems, such as subjective assessment, inaccurate targeting, and limitations in processing complex and multidimensional criteria data. This condition creates an urgent need for the implementation of an objective, accurate, and adaptive decision support system based on artificial intelligence. This study aims to design and implement an artificial intelligence-based Decision Support System (DSS) to determine the priority recipients of educational assistance in a well-targeted manner. The approach used is a hybrid method combining the Analytical Hierarchy Process (AHP) and Decision Tree. The AHP method is utilized to determine the importance weight of each assessment criterion, such as economic condition, academic achievement, number of dependents, and social status, based on a structured and consistent priority level. Furthermore, the Decision Tree algorithm is used to perform classification and make final decisions based on the weighted data patterns of assistance recipients.The research methodology includes the stages of data collection, determination of criteria and subcriteria, weight calculation using AHP, development of the Decision Tree classification model, and accuracy testing and validation of the system results. The system implementation is expected to produce recommendations for educational assistance recipients that are transparent, objective, and easy for policymakers to understand.

PENGUKURAN TINGKAT KEPUASAN PENGGUNA SIISTA STUDENT UNIVERSITAS DHYANA PURA MENGGUNAKAN METODE EUCS

2026-06-30T17:35:20+00:00

The Integrated Academic Information System (SIISTA) Student is used by students of Dhyana Pura University to access online academic services. However, users still encounter several issues, including unclear login error notifications, repeated login requests when accessing the Course Registration (KRS) feature, and unclear symbols in the tuition payment feature. This study aims to measure user satisfaction with SIISTA Student using the End User Computing Satisfaction (EUCS) method based on five variables: Content, Accuracy, Format, Ease of Use, and Timeliness. Data were collected from 97 active students through questionnaires. The results showed average scores of 3.95 for Content, 3.68 for Accuracy, 3.85 for Format, 3.68 for Ease of Use, and 3.59 for Timeliness. Overall, all variables were categorized as satisfied, although improvements are still needed in Accuracy, Ease of Use, and Timeliness to enhance system quality.

Deteksi Ekspresi Wajah Dalam Upaya Pencegahan Bullying Menggunakan Framework Tensor Flow dan Java Script Berbasis Website

2026-06-11T11:45:02+00:00

Facial expressions are changes in facial appearance that occur in response to a person's emotions, intentions, and social interactions. Facial expression analysis is an important research area because facial expressions serve as a non-verbal communication tool commonly used by humans to convey emotions, feelings, and social messages in everyday life. In many bullying cases, facial expressions may not accurately represent an individual's true emotional state. Some individuals may appear happy on the outside while internally experiencing sadness, anxiety, or intense emotional distress. Therefore, facial expression detection is needed to accurately identify emotional states and support bullying prevention efforts.

This study proposes a web-based facial expression detection system for bullying prevention using TensorFlow and JavaScript. The system is expected to provide accurate recognition of human emotions and feelings, enabling appropriate interventions and effective solutions for handling bullying cases. The research utilizes TensorFlow, an open-source machine learning library, along with JavaScript to recognize and classify seven basic human facial expressions, including happiness, sadness, anger, fear, surprise, disgust, and contempt, as well as a neutral expression. By accurately identifying emotional conditions, the proposed system is expected to contribute to more effective bullying prevention and support a safer social environment.