events

SEA: Search Engines Amsterdam

The 311th and 312th SEA Talks sessions bring two researchers to Amsterdam to discuss machine unlearning for retrieval and recommendation. Meghdad Kurmanji of IQVIA covers how unlearning is defined and evaluated, drawing on his work on SCRUB

Starts

2026-09-25T14:00:00+00:00

Organizer

SEA: Search Engines Amsterdam

Location

Amsterdam Science Park 904, Science Park 904, Amsterdam, Netherlands

The 311th and 312th SEA Talks sessions bring two researchers to Amsterdam to discuss machine unlearning for retrieval and recommendation. Meghdad Kurmanji of IQVIA covers how unlearning is defined and evaluated, drawing on his work on SCRUB, and asks what regulated deployments need beyond benchmark results. Pierre Sicco Lubitzsch of TU Berlin presents ERASE, a large-scale benchmark for unlearning in recommender systems spanning collaborative filtering, session-based and next-basket recommendation. The session runs at Science Park 904 with a parallel Zoom stream, and is aimed at researchers and practitioners working on search, recommendation and privacy compliance.