Decision Support System for Furniture Product Recommendation Using Item-Based Collaborative Filtering and Min-Max Normalization

Abstract

The increasing variety of furniture products available in the market makes it difficult for consumers to identify products that best match their preferences. Conventional product selection is often carried out manually, requiring consumers to compare product specifications one by one, which is time-consuming and may lead to less appropriate purchasing decisions. This study proposes a web-based decision support system for furniture product recommendation by integrating the item-based collaborative filtering (IBCF) method with min–max normalization. The recommendation process utilizes historical user ratings to identify similarities among products, while min–max normalization is applied to standardize product attributes, including price, category, material, color, and size, into a comparable scale. To address the cold-start problem for newly added products with insufficient rating data, a content-based similarity mechanism is incorporated into the recommendation process. The system was developed using a dataset consisting of 322 furniture products, 50 consumers, and 1,099 rating transactions. System functionality was evaluated using black-box testing, while user acceptance was assessed through user acceptance testing (UAT) involving ten respondents. The evaluation results show that all primary system functions operated as expected and achieved an overall UAT score of 88.0%, indicating that the proposed system is acceptable and capable of assisting consumers in selecting furniture products based on their preferences. This study contributes by integrating collaborative filtering, multi-attribute normalization, and a content-based cold-start mechanism into a single recommendation framework for furniture product recommendation