Group memory with feedback improves e-commerce recommendations for low purchasing power users
RecGroupMemory: Feedback-Aware Cohort Memory for E-Commerce Recommendation
Artificial Intelligence
Summary
E-commerce sites want to recommend products better by using group information about users with similar buying habits. The authors present a method called RecGroupMemory that updates these group profiles daily using the latest feedback to improve recommendations without losing important past information. Testing with users on Taobao who spend less showed that this feedback-aware approach increased the number of users who made transactions more than methods without memory or without feedback. This method helps recommendation systems learn and adapt more effectively over time.
What this means in practice
- •For e-commerce platform teams: Improve product recommendation accuracy for low-income user groups by incorporating daily feedback to update group profiles without losing prior observations.
- •For online marketing teams: Increase user purchase rates by using memory-augmented recommendation systems that revise guidance based on recent outcome feedback.
Authors
Wendong Zhang, Weipeng Huang, Zhixin Ma, Bin Zhang, Dimin Wang
Abstract
Cohort memory can support LLM-based recommendation planning through group-level user profiles and advice. As outcome feedback becomes available, this guidance may need to be updated while preserving the underlying observations. We present RecGroupMemory, a workflow that provides this memory to an orchestration LLM and separates memory construction from feedback-based revision. It uses feedback from the previous day to revise decision guidance, while validation checks keep selected source values unchanged. Online comparisons were conducted among Taobao users with low purchasing power selected by the business team. Relative to a no-memory baseline, memory without feedback and memory with feedback showed increases of 1.58% and 1.78%, respectively, in the number of guided transaction users.