Journal of Information Technology

Recommendations for Determining Tailoring Services at Ideal Tailor using the SMART Method

Fadhil Hakim — 2024-03-26

The sewing industry is one of the attractive sectors, which provides a variety of clothing, due to the growing interest in fashion. The demand for clothes with various models and styles is in great demand by various groups. The problem that often occurs is that many customers are confused when determining the model of material to be used. To assess the effectiveness of the web assessment system designed to help determine employee ratings at Ideal Tailor, which has been selected using the SMART method. In order to be able to help employees at Ideal Tailor in the assessment process which is very useful to get recommendations for customers who are in need of sewing services and help sewing shops in getting price, quality, service, and processing time ratings. Determination of recommendations for tailoring services will be carried out using this technique. The purpose of this research is to develop a decision support system application that can provide recommendations in choosing tailoring services provided by Ideal Tailor. The results of this programme show that the system built can help Ideal Tailor in determining recommendations for tailoring services at Ideal Tailor.

Design of a Decision Letter Management Information System (SISKA) at the SWCU Teacher Training and Education Faculty

Tri Ayu Nofitasari — 2024-03-26

Managing and archiving Decision Letters (SK) within the FKIP environment requires a management information system to make it more effective and efficient. This research aims to design a Decision Letter Management Information System (SISKA) that can make archiving easier and speed up searching and accessing SK archives. This type of research is qualitative, where the researcher developed the SISKA Web using the waterfall method. SISKA website uses PHP and MySQL. The Result of this research show that Manajemen Information System of Decision Letters based on web service can optimalize of managing and archiving Decision Letters within the FKIP UKSW. The SISKA website can monitor the process of proposing, creating, and disposing of Dean’s Decision Letters and Chancellor’s Decision Letters. Establishing the SISKA website, which is integrated with the FKIP UKSW’s website which it based on PHP, can improve the management and archiving of SKs at FKIP UKSW.

The Implementation of AHP and SAW Methods in Decision Support System to Determine BOS Fund Allocation

Gina Sania — 2024-03-26

SMP N 7 Padang Sidempuan encountered challenges in optimally allocating the School Operational Assistance Fund (BOS). The school's administration faced difficulties in determining the prioritization of fund usage during each period. The suboptimal utilization of the BOS could impede the achievement of national educational objectives and have adverse effects on human resource capabilities. This study combined the
AHP and SAW algorithms to determine the priority allocation of the School Operational Assistance Fund (BOS). The objective was to develop a web-based Decision Support System for budget management at SMP Negeri 7 Padang Sidempuan. With this approach, a system capable of determining the priority sequence
for BOS allocation based on its components was established. The research resulted in an integrated decision support system in the form of a website. This system effectively determined the priority sequence for BOS allocation based on various relevant components. With the implementation of this system, SMP N 7 Padang Sidempuan successfully managed the BOS, thereby maximizing its impact on improving the quality of education.

Implementation of Neighborhood-Based Collaborative Filtering Algorithm on Laundry Service Recommendation System

Tony Darmanto — 2024-03-26

Modern information technology has advanced human life significantly, particularly in the way that it presents crucial information. People often find it challenging to choose the right kind of information for their needs due to the sheer volume of data that information technology may offer. To assist users in selecting the information they require, a recommendation system has been implemented. It is possible to use collaborative filtering algorithms to the use of recommendation systems in industries like laundry services. by the application of an experimental, causal connection research design as the research methodology. A literature review, which comprises scientific books, research reports, scientific journals, theses, and textual materials in both print and electronic format, is the method of data collection employed by the author. The Unified Modeling Language (UML), which the author uses to describe the processes by which the collaborative filtering recommendation system applies its recommendations for laundry services, is one of the methods of system modeling that the author employs. The collaborative filtering algorithm creates new information based on the pattern of a group of users who share similar features by filtering laundry service data according to the user's desired attributes. The filtering results are displayed through an interactive user interface in an algorithm that is implemented in a web application. The user's demands and preferences can be taken into account when generating suggestion results by the laundry service recommendation system.

Scheduling Inmate Learning: Implementation of Genetic Algorithm at LPKA Class I Tanjung Gusta Medan

Yun Betry Siagian — 2024-03-27

Scheduling in education has been a common issue faced, particularly concerning the mismatch between class hours and teacher availability. Addressing this necessitated careful precision and a significant amount of time for subject scheduling. Genetic algorithms emerged as a suitable choice for managing school subject scheduling due to their capability to address multi-criteria and multi-objective problems inspired by biological evolution. The concept of genetic algorithms could be effectively applied in the context of schedule arrangement within educational institutions. This study employed a quantitative methodology and implemented the Rapid Application Development (RAD) system development method. The utilization of genetic algorithms in subject scheduling, as demonstrated in this research, was deemed beneficial for correctional institutions previously engaged in manual scheduling processes. The outcomes indicated the success of the genetic algorithm approach in resolving school scheduling issues by seeking optimal scheduling combinations, as evidenced by achieving maximum fitness values and minimal errors (in this case, an error value of 0), indicating the absence of class schedule conflicts.

Recognition Of Kalbar Dunging Script Patterns using The Learning Vector Quantization (Lvq) Method

Maya Sari — 2024-03-27

This article implements the Learning Vector Quantization (LVQ) method in recognizing Dunging character patterns. In this case, it is a system that groups dunging characters according to their respective classes. By applying these two theories, the system will identify handwritten dunging script which has characteristics resembling human neural networks. The data used is in the form of images that have been taken using digital camera photos and smartphones.
The image is converted into numeric using image processing. The image processing stages include the process of cropping the RGB image to 50 x 50 pixels, and image binarization. The network training stage uses data of 10 images consisting of 10 characters of Dunging script. The results of testing the training image obtained a percentage of 88.66%. Using dunging child diacritics, using or adding some image processing functions, and implemented. From the accuracy results obtained, it can be said that the LVQ method is not optimal in solving pattern recognition problems, especially Dunging characters. Optimization techniques for the LVQ learning process with optimization algorithms are the next research plan.

Hybrid Naive Bayes Classifier and Markov Chain for Predicting the Level of Quick Response Indonesian Standard (QRIS) Usage

Chairina Chairina — 2024-03-27

Digital based payment systems have provided a variety of payment formats and interactions. Considering the benefits and effectiveness of the Quick Response code, Bank Indonesia has designated this technology as the standard for using payment methods. The Hybrid Naϊve Bayes Classifier and Markov Chain methods are classification and prediction methods that can be used to predict the level of QRIS use in this research. The data used comes from Bank Indonesia data for Medan City from January 2020 to February 2023 with 3 parameters for the level of QRIS usage. The research results show that with testing data the classification results of the level of use of QRIS as an electronic payment tool in Indonesia have increased and only in December 2022 has it decreased. Then, based on the percentage probability calculation value using the Markov Chain method with n going to infinity, the probability value in the transition probability matrix converges to a value. This value is the same as the opportunity value calculated by limiting probability in the Markov Chain. The percentage chance value for the level of use of QRIS as an electronic payment tool in Indonesia has an increase of 0.7076 or 70.76%, and a percentage chance of a decrease of 29.24 or 29.24%.

Application of Naïve Bayes Method for Diagnosing Rubber Tree Diseases at PT. Bridgestone Aek Tarum

Efriliya Hafni Yuswinda — 2024-03-27

Rubber plant disease attacks roots, stems, tapping areas, and leaves. Factors causing the low rubber plantations are due to limited information and control of rubber tree diseases, so there is a need for a system to improve rubber tree disease control with the help of this system. Currently, without a system, it is only completed by an expert who is considered successful if he is able to make decisions like an expert. Making this system can help an employee in diagnosing diseases in rubber plants. Making this system requires an expert in his field to obtain accurate data regarding information on rubber plant diseases, the data used are 22 symptoms and 7 diseases. In this study the method used is Naive Bayes. The Naive Bayes method works by finding the probability value of the emergence of rubber plant diseases. This application is developed based on Android using the Java programming language in the Android Studio application. From this study it was found that white root fungal disease has 7 symptoms and the choice of confidence level in each symptom is between 0 and 1. Then white root fungal disease has the largest result value, namely 0.0000002911.