Publication Search

This research aims to develop a multiclass flower image classification system using the Convolutional Neural Network (CNN) algorithm with the EfficientNet architecture. The main problem addressed is the difficulty of manual identification of flower species that share high visual similarity. The research stages include collecting 17,299 flower images across 19 classes, performing data preprocessing such as image resizing, pixel normalization, and augmentation, followed by model training using the EfficientNet transfer learning approach. The model was trained for 10 epochs with an 80:20 training-validation data split. The evaluation results show that the model achieved a validation accuracy of 98.05% with a loss value of 0.0968, and an average precision, recall, and F1-score of 0.98. The trained model was then implemented into a web-based application built using the Next.js framework, enabling users to upload flower images and obtain real-time classification results via the Hugging Face API. The system successfully identified flower species with a confidence level of 99.87%. These findings demonstrate that combining a modern CNN architecture with transfer learning provides efficient and highly accurate flower classification performance, which can be effectively implemented for educational and digital conservation purposes.

This study presents the integration of a Large Language Model (LLM) Ollama with the OpenStreetMap (OSM) API within a Business Intelligence (BI) framework to develop an intelligent, location-based recommendation system. The system is designed to assist users in finding dining, leisure, and resting places through natural language interaction and contextual understanding. The LLM interprets user input semantically, transforms it into structured spatial queries, and retrieves relevant geospatial data from OSM. The data are then analyzed, categorized, and visualized using BI methods to enhance interpretability and decision-making. The system was implemented using Next.js, Leaflet.js, ensuring interactivity and scalability for web-based deployment. Technical evaluation focused on system accuracy, response time, and output consistency. Results demonstrate an average response time of 1.74 seconds, 80% accuracy, and 80% consistency, proving the model’s efficiency in producing relevant, context-aware recommendations. This integration highlights the potential of combining open geospatial data, local LLMs, and BI analytics to create intelligent, data-driven decision support systems applicable to tourism, urban planning, and spatial information management.

This study is dedicated to the design and implementation of a web-based inventory management platform for PT. Paramata Baraya International. replacing the manual Excel system that is ineffective and susceptible to mistakes. The Rapid Application Development (RAD) method is chosen for its quick and iterative approach to software development. The system is built using the Next.js framework and PostgreSQL as the database, with Prisma ORM for database management. Key features include item management, real-time status tracking, calibration, maintenance, equipment rental, and automatic notifications. The implementation results show a 60% increase in item search efficiency, 95% data accuracy, and an 80% reduction in manual input errors. This study is significant as a solution for companies engaged in calibration and maintenance of measuring instruments.

This research aims to develop the company website PT. Rantangin Digital Indonesia uses the latest technology, namely Next.js and Tailwind CSS. This website development was carried out to increase visibility and user experience in accessing the services offered by the company. Next.js was chosen as the development framework because of its ability to build fast, scalable and SEO-friendly web applications, while Tailwind CSS was used to speed up the interface design process with a flexible and responsive utility-first approach. The development method used is the Waterfall method, which consists of planning, design, implementation, testing and maintenance stages. The development results show that the website built has succeeded in fulfilling the company's goals of providing a digital platform that is easy to access, responsive on various devices, and has optimal performance. By using Next.js and Tailwind CSS, website development can be done efficiently and effectively, and provide a better user experience. Recommendations for further development include integration with content management systems and analytical features for monitoring website performance.

The development of information and communication technology in Indonesia has grown significantly over the past few years. Thus, there are many opportunities to digitize the business process of a business or business ecosystem. One of them is in the field of digital economy by adapting business processes into a digital platform, such as designing an e-commerce platform in a village that has commodities in it so that related commodities can increase the visibility of their products. This research uses Belega Village as a case study location considering that this village has a commodity in the form of bamboo crafts. Research was conducted to build a web-based bamboo handicraft e-commerce application that would be expected to increase product visibility. The application itself is built using the Agile Development method using the sprint model. This application is built with a microservice architecture with Express.js technology as a backend service and Next.js to create client applications and PostgreSQL as a DBMS. The result of this research is a web-based e-commerce application that is expected to increase the visibility of products from bamboo crafts from Belega Village.