New Perspectives on Query Performance Prediction
Abstract
Query Performance Prediction (QPP) is the task of estimating retrieval effectiveness in the absence of relevance judgments. This thesis presents new perspectives on QPP evaluation and methodology, introducing novel frameworks including ANOVA-based modeling, query variation analysis, and user-centric approaches to performance prediction.
Type
Publication
PhD Thesis, RMIT University
Supervisors: Prof. J. Shane Culpepper and Prof. Falk Scholer
Institution: RMIT University, Melbourne, Australia
Year: 2024