Energy & materialsResearch findings
Development of a Behavior-Based Recommendation System
Original published 2024.03.27 22:12 KSTAdded to Blackdeck 2026.10.27 19:06 KST
What has changed?
A new method leveraging browsing and clicking behaviors to generate personalized recommendations is proposed
Key points
- Existing rating-based systems face limitations due to sparsity and lack of collaboration; this study addresses these by using browsing and点击
- An unsupervised clustering method based on product categories enables more accurate grouping of users with similar behaviors
- Neighborhoods of active customers within clusters are defined to recommend high-reputation products with higher relevance
Connect it to your work
The system improves recommendation accuracy by analyzing natural user behaviors on e-commerce platforms
What remains unproven
This research note is based on an abstract; check the original for test conditions and application scope.
Originals used for this note
Links open the original publisher in a new tab.