Pengembangan Pencarian Visual Berbasis CNN Dalam Manajemen Platform Jual Beli Online Produk UMKM

Oktaviano, Kevin Harlis (2024) Pengembangan Pencarian Visual Berbasis CNN Dalam Manajemen Platform Jual Beli Online Produk UMKM. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6047222004-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (5MB) | Request a copy

Abstract

Usaha Mikro, Kecil, dan Menengah (UMKM) memainkan peran penting dalam perekonomian Indonesia. Namun, UMKM seringkali menghadapi tantangan dalam hal manajemen produk dan interaksi dengan pelanggan. Penelitian ini bertujuan untuk mengembangkan platform pencarian visual berbasis Deep Learning untuk identifikasi produk UMKM guna mengatasi tantangan tersebut. Metode penelitian diawali dengan studi literatur tentang Deep Learning dan CNN (Convolutional Neural Network). Selanjutnya dilakukan pengumpulan data berupa gambar produk UMKM, diikuti pra-pemrosesan dan augmentasi data. Kemudian dirancang arsitektur CNN VGG16 yang terdiri dari 16 lapisan untuk feature extraction dan classification. Model terbaik yang memenuhi target performa kemudian diintegrasikan pada prototipe platform berbasis web. Platform pencarian visual ini diharapkan dapat membantu pengelolaan produk UMKM serta meningkatkan pengalaman pelanggan dalam mencari dan menemukan produk yang mereka butuhkan. Secara keseluruhan, penelitian ini bertujuan mendukung pertumbuhan UMKM di Indonesia melalui adopsi teknologi Deep Learning.
=================================================================================================================================
Ministry of Micro, Small and Medium Enterprises (MSMEs) play a big role in Indonesia's economy. However, MSMEs often face challenges in managing their products and engaging with customers. This research aims to develop a visual search platform using Deep Learning to identify MSME products to address these challenges. The research methodology starts with studying Deep Learning and Convolutional Neural Networks (CNNs) from published literature. Next, images of MSME products are collected, pre-processed and augmented to create a robust dataset. An architecture called VGG16 CNN with 16 layers is then designed to extract visual features from product images and classify them. The best performing model is integrated into a web-based prototype platform. The visual search platform is expected to assist MSMEs in better product management and improve customer experience in searching for and discovering the products they need. Overall, this research aims to support MSME growth in Indonesia through the adoption of Deep Learning technology. In everyday language, this platform should make it easier for small businesses to organize and sell their products online, and help buyers find the right products more easily using image search.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Deep Learning, Convolutional Neural Network, Visual Search, E-commerce, Produk UMKM
Subjects:	T Technology > T Technology (General)
Divisions:	Interdisciplinary School of Management and Technology (SIMT) > 78201-System And Technology Innovation
Depositing User:	Kevin Harlis Oktaviano
Date Deposited:	28 Aug 2024 03:08
Last Modified:	28 Aug 2024 03:08
URI:	http://repository.its.ac.id/id/eprint/115229

Actions (login required)

View Item