Jurnal Teknologi dan Sistem Komputer
Not a member yet
364 research outputs found
Sort by
Sistem Pakar Untuk Identifikasi Dan Alternatif Solusi Terhadap Permasalahan Yang Dihadapi Peserta Didik Sekolah Menengah Menggunakan Rule-Based Machine Learning
Full text linkKesulitan dalam mengidentifikasi masalah yang dialami oleh peserta didik di sekolah tingkat menengah berpotensi mengakibatkan masalah yang lebih besar di kemudian hari. Selama ini guru bimbingan dan konseling (BK) menggunakan metode konvensional dalam mengidentifikasi dan memecahkan masalah tersebut. Metode ini membutuhkan biaya yang besar, ruang yang khusus, dan waktu yang lama. Artikel ini memaparkan pengembangan sistem pakar untuk identifikasi untuk kemudian menawarkan alternatif solusi terhadap permasalahan yang dihadapi peserta didik di tingkat sekolah menengah. Sistem ini menggunakan metode Problem Checklist yang didukung oleh machine learning untuk meningkatkan akurasi dan efisiensi. Kepakaran dari guru BK senior untuk menghubungkan ragam permasalahan dan berbagai alternatif solusi dilatihkan pada rule-based machine learning sistem ini. Pengujian dilakukan menggunakan WEKA dengan 200 contoh data sampel sebagai data pelatihan, yang dapat memprediksi data dari 185 contoh yang label kelasnya tidak diketahui Sistem yang dikembangkan adalah berbasis web sehingga peserta didik yang merasa mengalami masalah dapat meng-ases dirinya sendiri dan mendapatkan saran alternatif solusi secara online. Guru BK sebagai admin dapat memantau perkembangan peserta didiknya serta melakukan penambahan ragam permasalahan dan alternatif solusi mengikuti perkembangan jaman. Hasil implementasi dan pengujian menunjukkan bahwa sistem pakar yang dikembangkan menawarkan identifikasi dan solusi yang akurat dan lebih cepat serta dapat dilakukan kapanpun dan di manapun
Pemanfaatan Konfigurasi Layer Pada Metode CNN Untuk Peningkatan Kinerja Klasifikasi Penyakit Daun Tomat
No full textTomat adalah salah satu komoditas hortikultura dengan nilai ekonomi yang tinggi, tantang yang dihadapi oleh petani salah satunya dalah kerentanan penyakit tomat terhadap penyakit. Identifikasi secara visual pada daun sulit diuraikan dengan sekali pandang, sehingga menyebabkan asumsi yang tidak akurat tentang penyakit tersebut. Akibatnya, mekanisme pencegahan yang dilakukan petani menjadi tidak efektif dan berdampak merugikan. Penelitian ini mengusulkan identifikasi penyakit tomat secara automatis menggunakan metode Convolution Neural Network. Dalam makalah ini kami melakukan evaluasi pada metode CNN dengan arsitektur Alexnet dengan konfigurasi layer untuk mencari hasil kinerja terbaik dari penggunaan parameter tersebut pada architektur Alexnet. Pada penelitian ini juga melakukan analisis yang diperoleh dari hubungan antara parameter yang digunakan terhadap kinerja akurasi, dan analisis terhadap dampak penggunaan parameter dengan jumlah dataset daun tomat dari dataset PlantVillage
Evaluation of Decentralized Website Performance Using Blockchain DNS
Full text linkPrivacy violations and misuse of customer data are the failures of Web 2.0. Web 2.0 employs a centralized server that can be accessed by authorities. With the development of Web 3.0, which is built on blockchain and uses a decentralized application (dApp) to avoid privacy violations and data misuse. Decentralized systems can offer users privacy and autonomous control of the system, which is an important feature of consumer protection. Decentralized web development, also known as Web 3.0, was created using blockchain technology and the peer-to-peer network protocol (IPFS). In a decentralized website environment, data on the server can be linked to conventional domains (.com, .id, etc.) or to special domains provided by blockchain-based platform domains. Every transaction must be linked to the Ethereum wallet and cryptocurrency for the smart contract to function. This activity takes a long time and is a bad UX for decentralized web developers. To compare the performance of a decentralized web using blockchain DNS and a centralized web using conventional domains, a 15-minute usage scenario with a maximum of 50 users is used. The results show that the throughput and bandwidth of a decentralized web is higher than a centralized web
Aspect-Based Analysis of Telkomsel User Sentiment on Twitter Using the Random Forest Classification Method and Glove Feature Expansion
Full text linkIn this modern era, people certainly very easy to access social media, one of which is Twitter. Twitter is usually used by the public in expressing opinions regarding current issues, product reviews, and many other things positive, negative, or neutral opinions, or can be interpreted as sentiment. This study aims to analyze the aspect-based sentiment of Telkomsel users on Twitter using random forest classification and the extension of the Glove feature. This study uses signal aspects and service aspects with a total dataset of 16988 data. A Random forest can be classified as relevant and accurate for sentiment analysis with the greatest accuracy of 80.37% in the signal aspect and 80.12% in the service aspect, and the expansion feature is proven to be able to increase the performance value of this study by 13.15% in the signal aspect. and 5.37% in the service aspect
Large-scale integrated infrastructure for asynchronous microservices architecture
Full text linkIntegrated large-scale business activities increasingly rely on the use of remote resources and services across multi-platform applications. Microservice in previous research has become a solution, but this approach still leaves a data loss problem. This research methodology proposed an architecture of data transmission managed by messaging service to prevent data loss in handling many requests to deliver a multiplatform architecture, handling the plugin services, and enabling escalation based on the requirement. As a result, this research successfully implements large-scale multiplatform Single Sign-On (SSO) infrastructure for asynchronous microservices architecture. The system test results show that the developed system can handle up to 2000 requests with 20 concurrent requests
Comparison of various epidemic models on the COVID-19 outbreak in Indonesia
Full text linkThis paper compares four mathematical models to describe Indonesia's current coronavirus disease 2019 (COVID-19) pandemic. The daily confirmed case data are used to develop the four models: Logistic, Richards, SIR, and SEIR. A least-square fitting computes each parameter to the available confirmed cases data. We conducted parameterization and sensitivity experiments by varying the length of the data from 60 until 300 days of transmission. All models are susceptible to the epidemic data. Though the correlations between the models and the data are pretty good (>90%), all models still show a poor performance (RMSE>18%). In this study case, Richards model is superior to other models from the highest projection of the positive cases of COVID-19 in Indonesia. At the same time, others underestimate the outbreak and estimate too early decreasing phase. Richards model predicts that the pandemic remains high for a long time, while others project the pandemic will finish much earlier
TATOPSIS: A decision support system for selecting a major in university with a two-way approach and TOPSIS
Full text linkSeveral problems can occur when students feel they have made the wrong choice of major in university. Choosing a major is one of the problems that students often face. Therefore, this study aims to develop a Decision Support System (DSS) to help students find majors that match their interests and abilities. This DSS proposes a two-way approach by considering students and the major's requirements, standards, and characteristics. The DSS utilizes the TOPSIS method; therefore, it is called TATOPSIS, which stands for Two-way Approach TOPSIS. It showed that the two-way approach in Scenario 1 (without score normalization) and Scenario 3 (with score normalization) shows better agreement results in 78.33% and 73.33% than the two-way approach for Scenario 2, Scenario 4, and the student-one-way approaches
Data scaling performance on various machine learning algorithms to identify abalone sex
Full text linkThis study aims to analyze the performance of machine learning algorithms with the data scaling process to show the method's effectiveness. It uses min-max (normalization) and zero-mean (standardization) data scaling techniques in the abalone dataset. The stages carried out in this study included data normalization on the data of abalone physical measurement features. The model evaluation was carried out using k-fold cross-validation with the number of k-fold 10. Abalone datasets were normalized in machine learning algorithms: Random Forest, Naïve Bayesian, Decision Tree, and SVM (RBF kernels and linear kernels). The eight features of the abalone dataset show that machine learning algorithms did not too influence data scaling. There is an increase in the performance of SVM, while Random Forest decreases when the abalone dataset is applied to data scaling. Random Forest has the highest average balanced accuracy (74.87%) without data scaling
Sistem Penghitung Jumlah Orang Menggunakan Metode SSD-MobileNet dan Centroid Tracking
Full text linkSalah satu penerapan kecerdasan buatan untuk mencegah penyebaran virus corona adalah dengan membuat sistem penghitung jumlah orang otomatis untuk mencegah kerumunan di dalam ruangan. Penelitian ini membahas mengenai pembuatan prototipe sistem penghitung jumlah orang menggunakan algoritma deep learning pada single board computer. Tujuan dari penelitian ini adalah untuk menghitung jumlah orang dalam suatu ruangan agar okupansi ruangan dapat ditekan. Kontribusi dari penelitian ini adalah mengkombinasikan dua metode visi komputer yaitu SSD-MobileNet untuk identifikasi objek orang dan centroid tracking untuk menghitung jumlah orang. Berdasarkan pengujian yang telah dilakukan menunjukan bahwa sistem telah dapat menghitung objek orang dengan akurasi 100% apabila jumlah orang yang memasuki ruangan berjumlah satu, dua, atau tiga secara bersama-sama. Kemudian sistem dapat mendeteksi objek dengan jarak maksimal 10 meter dan intensitas cahaya redup atau kurang dari 100 lux. Pada pengujian komputasi menunjukan bahwa sistem dapat memproses video dengan jumlah frame 30 fps dan kualitas video high definition (HD)
Spatial Skyline Query Based on Surrounding Environment Untuk Data Streaming Menggunakan Apache-Spark
Full text linkPrevious research on Spatial Skyline Query Based on Surrounding Environment left a challenge in finding skyline objects that support the use of mobile devices. This study introduces a method that allows users to search for spatial objects dynamically. Cloud-based streaming data services are currently available to support the dynamic search of spatial skyline objects. Under these conditions, streaming data requires a longer processing time. This study aims to examine the effectiveness and efficiency of Apache-Spark in developing Spatial Skyline Query Based on Surrounding Environment in processing streaming data. Further implementation of the developed algorithm can provide better location access for users on mobile devices. Comparative analysis of algorithm execution time is performed by comparing algorithm processing on a single processor and cluster computing using various evaluation parameters. The test results on each parameter show that the computation time of the proposed algorithm on a single computation is not as good as the previous algorithm. However, in cluster computing, the proposed algorithm is superio