ITEM SELECTION RECOMMENDATION FOR FASHION’S E-COMMERCE BASED ON CONSUMER’S REVIEW

• Main Author: Nur Huzaifah, Ridha
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/65202
• Institution: Institut Teknologi Bandung

E-commerce technology has developed rapidly in meeting the needs of community. After the pandemic Covid-19 and community activity restrictions, the use of e-commerce among the public has increased for fulfilling their daily needs or desires. Certainly, finding the desired item on e-commerce is not an easy thing and requires a lot of time. One of the things that causes this problem is the larger number of goods provided by the e-commerce. To overcome this issues, users usually will ask for recommendations from their friends, relatives, or consider the recommendations given by e-commerce itself to find items they are looking for. Currently, e-commerce companies are competing to display the best item recommendations to increase user’s purchasing. Some methods will be definitely required to produce the best item recommendations. In this final project, two recommendation system approaches will be applied, namely non personalized and personalized recommendation system. For the non personalized recommendation, the average rating technique and weighted rating will be used. While content-based filtering and collaborative filtering will be used in a personalization-based approach. To produce the best recommendations, a hybrid recommendation system will be used to cover the shortcomings of a recommendation system with other recommendation system. Based on the simulation, the hybrid recommendation system is the most suitable recommendation system to use. The recommended clothing items from this recommendation system are items with clothing ID 1035, 918, 745, 1099, and 107.