Sistem Rekomendasi Dengan Teknik Collaborative Filtering Menggunakan Metode Faktorisasi Matriks Dan K-Nearest Neighbor (Studi Kasus: Produk Elektronik Pada E-Commerce)

Yafie, Haris (2023) Sistem Rekomendasi Dengan Teknik Collaborative Filtering Menggunakan Metode Faktorisasi Matriks Dan K-Nearest Neighbor (Studi Kasus: Produk Elektronik Pada E-Commerce). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06211940000074-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2025. Download (2MB) | Request a copy

Abstract

Perkembangan teknologi menghasilkan e-commerce yang telah menghadirkan banyak pilihan produk elektronik, yang sering membuat user e-commerce bingung dalam memilih produk. Untuk mengatasi hal ini, diperlukan sistem rekomendasi yang menggunakan metode collaborative filtering. Penelitian ini menggunakan metode faktorisasi matriks, khususnya singular value decomposition (SVD), dan metode K-Nearest Neighbor (KNN). Tujuan penelitian ini adalah mengembangkan sistem rekomendasi menggunakan collaborative filtering dengan faktorisasi matriks dan KNN untuk merekomendasikan produk elektronik di e-commerce. Penelitian ini diharapkan dapat membantu user dalam memilih produk, meningkatkan penjualan dan loyalitas user e-commerce, serta menjadi referensi untuk penelitian selanjutnya di bidang sistem rekomendasi di e-commerce. Data rating user terhadap produk elektronik diperoleh dari University of California San Diego (UCSD). Sistem rekomendasi faktorisasi matriks SVD memiliki performa lebih baik dalam memprediksi rating dan merekomendasikan produk elektronik daripada sistem rekomendasi KNN. Sistem rekomendasi KNN memiliki kelemahan dalam mengolah data yang besar sedangkan faktorisasi matriks dapat mengatasi hal tersebut. Saran yang diberikan peneliti adalah menggunakan data produk lainnya dan penambahan variabel untuk meningkatkan akurasi rekomendasi. Mengingat keterbatasan memori pada metode KNN, penelitian selanjutnya disarankan menggunakan metode lain seperti content based filtering dan hybrid filtering untuk hasil yang lebih baik dan inovatif.
======================================================================================================================================
Technological advancements have led to the emergence of e-commerce, offering a wide range of electronic products that often leave users perplexed when making choices. To address this issue, a recommendation system utilizing collaborative filtering methods is required. This research adopts the matrix factorization technique, particularly singular value decomposition (SVD), along with the K-Nearest Neighbor (KNN) algorithm. The objective is to develop a collaborative filtering recommendation system employing matrix factorization and KNN to suggest electronic products in e-commerce. The anticipated outcomes include assisting users in product selection, enhancing sales and user loyalty in e-commerce, and serving as a reference for future research in the field of recommendation systems. User ratings for electronic products were collected from the University of California San Diego (UCSD). The SVD matrix factorization system outperforms the KNN system in terms of rating prediction and recommending items. KNN system has a weakness which is processing using large dataset while the SVD matrix factorization can solve that. Researchers suggest utilizing different datasets and additional variables to enhance recommendation accuracy. Due to the memory limitations of the KNN algorithm, exploring alternative methods like content-based filtering and hybrid filtering is recommended for more innovative and improved outcomes.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	e-commerce, faktorisasi matriks, k-nearest neighbor, sistem rekomendasi, e-commerce, k-nearest neighbor, matrix factorization, recommendation system.
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Q Science > QA Mathematics > QA76.9.I58 Recommender systems (Information filtering)
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Haris Yafie
Date Deposited:	10 Aug 2023 03:38
Last Modified:	10 Aug 2023 03:38
URI:	http://repository.its.ac.id/id/eprint/104373

Actions (login required)

View Item