Aditya Wicaksono, Setiady Ibrahim Anwar, Muhammad Al Amin, Rizki Juliansyah, Arya Dimas Wicaksana, Gema Parasti Mindara

E-Commerce FreezeMart untuk Penjualan Frozen Food dengan Sistem Rekomendasi Berbasis Content-Based Filtering

Introduction

E-commerce freezemart untuk penjualan frozen food dengan sistem rekomendasi berbasis content-based filtering. FreezeMart: E-commerce makanan beku dengan sistem rekomendasi content-based filtering dan NLP. Personalisasi pengalaman belanja, kelola pesanan, dan dapatkan saran produk relevan.

Abstract

Kemajuan teknologi digital mendorong perkembangan platform e-commerce yang memudahkan transaksi secara cepat dan fleksibel. Penelitian ini membahas FreezeMart, situs e-commerce untuk penjualan makanan beku yang dilengkapi sistem rekomendasi terpersonalisasi. Sistem memanfaatkan content-based filtering dan analisis histori pembelian untuk menyarankan produk yang relevan. Masukan pengguna diproses menggunakan teknik Natural Language Processing seperti TF-IDF untuk mencocokkan preferensi dengan atribut produk. Pengembangan dilakukan dengan metode Agile Scrum dan mencakup tiga peran utama: guest, customer, dan admin. Pengguna dapat mengelola pesanan dan memperoleh rekomendasi sesuai kebutuhan, sementara admin bertanggung jawab atas pengelolaan sistem. Sistem rekomendasi yang diimplementasikan terbukti mampu meningkatkan pengalaman belanja pengguna. Pada tahap selanjutnya, pengembangan sistem dapat difokuskan pada penerapan collaborative filtering serta peningkatan kinerja platform.

Review

This paper introduces FreezeMart, an e-commerce platform specifically designed for the sale of frozen food, featuring a personalized recommendation system. The core innovation lies in its application of content-based filtering, which intelligently leverages user purchase history and natural language processing (NLP) techniques, such as TF-IDF, to interpret user preferences and suggest relevant products. The project's development, utilizing the Agile Scrum methodology, addresses a specific market need by combining specialized product offerings with advanced personalization, a commendable approach in the rapidly evolving digital commerce landscape. The technical implementation of FreezeMart's recommendation engine is well-outlined, focusing on a content-based approach that effectively matches product attributes with user interests. The inclusion of TF-IDF for processing user input demonstrates a thoughtful consideration for accurately capturing and acting upon explicit user preferences, thereby enhancing the relevance of recommendations. The defined system roles—guest, customer, and admin—also indicate a structured design capable of managing diverse user interactions and administrative tasks, pointing to a robust and user-centric system architecture. While the abstract positively claims that the implemented recommendation system significantly improves the user shopping experience, the specifics regarding the empirical evidence or evaluation methodology supporting this assertion are notably absent. For future iterations, substantiating these claims with quantitative metrics, such as A/B testing results, conversion rates, or user satisfaction surveys, would greatly enhance the paper's academic rigor. The suggestions for future work, including the adoption of collaborative filtering to broaden recommendation diversity and a focus on overall platform performance enhancement, are pertinent and demonstrate a clear path for further development and research.

Full Text

Comments

You need to be logged in to post a comment.