DEVELOPMENT OF A CLOTHING AND FOOTWEAR GIFT RECOMMENDATION SYSTEM USING IMAGE- BASED CONVOLUTIONAL NEURAL NETWORKS (CNN) AND K-NEAREST NEIGHBORS (K-NN) ALGORITHM

Abstract

In an increasingly digital world, human relationships are often overshadowed by technology. However, giving gifts remains a fundamental means of strengthening social bonds. Choosing the perfect gift can be a difficult and time-consuming activity, especially when it comes to clothing and footwear, where everyone's fashion tastes vary greatly. Keeping these challenges in mind, this project presents a novel solution in the form of an image-based reward recommendation system that leverages the power of deep learning Convolutional Neural Networks (CNN) and K-Nearest Neighbors (K-NN) algorithms. Through Resnet50's deep learning techniques, the system identifies visual characteristics, while the K-NN algorithm matches user preferences with visually similar products in the dataset, thereby simplifying the reward selection process. This research aims to create a simple and efficient system that increases gift choices by reducing complexity, encouraging meaningful gift giving, and strengthening social relationships.