Curriculum Approximate Unlearning for Session-based Recommendation 452 CCS: Information systems Recommender systems CCS: Security and privacy Web application security

Liu Yang, Zhaochun Ren, Ziqi Zhao, Pengjie Ren, Zhumin Chen, Xinyi Li, Shuaiqiang Wang, Dawei Yin, Xin Xin · 2025

Read Curriculum Approximate Unlearning for Session-based Recommendation 452 CCS: Information systems Recommender systems CCS: Security and privacy Web application security by Liu Yang, Zhaochun Ren, Ziqi Zhao, Pengjie Ren, Zhumin Chen, Xinyi Li, Shuaiqiang Wang, Dawei Yin, Xin Xin (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-15263.txt.

Description

Liu Yang and co-authors present a framework for making recommendation systems forget chosen training data, in a 2025 paper. Session based recommenders learn from user activity, and removing the influence of specific samples without full retraining is called approximate unlearning. The standard gradient ascent trick can damage recommendation quality.

Their framework, CAU, treats unlearning and retention as two goals to balance, seeking a compromise where neither suffers much. It also orders the samples to forget from easy to hard, a curriculum idea, using two new difficulty measures. That order improves both the forgetting effect and the surviving accuracy. The authors argue request order matters when many deletions arrive at once. The method targets privacy driven deletion in web services.

Catalog record

CREATORLiu Yang, Zhaochun Ren, Ziqi Zhao, Pengjie Ren, Zhumin Chen, Xinyi Li, Shuaiqiang Wang, Dawei Yin, Xin Xin
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE52 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-15263

Browse the full library · About the Tower of Babel