Report on the 2nd Workshop on Query Performance Prediction and its Applications in the Era of Large Language Models (QPP++ 2025) at ECIR 2025
Chuan Meng, Guglielmo Faggioli, Mohammad Aliannejadi, Nicola Ferro, Josiane Mothe · ACM SIGIR Forum · 2025
Query performance prediction (QPP) is a key and long-standing task in information retrieval (IR). The task of QPP aims to predict ranking quality without human relevance judgments. The rise of large language models (LLMs) has significantly reshaped the IR research landscape. These advancements call for a timely discussion on how we perform, evaluate and apply QPP in the LLM era. To foster this discussion, we have organised the workshop on QPP at the 47th European Conference on Information Retrieval (ECIR 2025). While LLM-related topics were a major focus, we made sure to keep our workshop open to a broad range of QPP-related research areas. The workshop featured five accepted papers and gathered around 20 participants for a half-day of presentations and interactive discussions. These discussions produced several key insights and directions for future QPP research. Date: 10 April 2025. Website: https://qppworkshop.github.io/.