<![CDATA[Disorganized domestic waste management has become a serious environmental issue due to the lack of systems for the rapid and automated recording of waste types. This study aims to design and develop a web service-based information system capable of automatically classifying domestic waste types. The research employs a Convolutional Neural Network (CNN) for waste image recognition, integrated with a web service architecture using a REST API to facilitate data exchange between systems. The dataset comprises thousands of domestic waste images categorized into major groups such as organic, plastic, paper, and metal. Key testing results demonstrate that the developed CNN model exhibits excellent and stable performance, achieving an accuracy rate of 89.03% and a loss value of 0.29 on the test data. The study concludes that integrating the CNN method with a web service is effective, accurate, and viable for automating domestic waste type identification, thereby supporting smarter environmental management systems.]]>
Copyrights © 2026
