Acknowledgments
Introduction - The Robustness Problem of Information Retrieval
Reasons for retrieval failures - the RIA workshop
Instability in retrieval - the TREC's Robust tracks
Estimating the query difficulty
Summary
Basic Concepts
The retrieval task
The prediction task
Linear correlation
Rank correlation
Prediction robustness
Summary
Query Performance Prediction Methods
Pre-Retrieval Prediction Methods
Linguistic approaches
Statistical approaches
Definitions
Specificity
Similarity
Coherency
Term relatedness
Evaluating pre-retrieval methods
Summary
Post-Retrieval Prediction Methods
Clarity
Definition
Examples
Other Clarity measures
Robustness
Query perturbation
Document perturbation
Retrieval perturbation
Cohesion
Score distribution analysis
Evaluating post-retrieval methods
Prediction sensitivity
Summary
Combining Predictors
Linear regression
Combining pre-retrieval predictors
Combining post-retrieval predictors
Combining predictors based on statistical decision theory
Evaluating the UEF framework
Results
Combining predictors in the UEF model
Summary
A General Model for Query Difficulty
Geometrical illustration
General model
Validating the general model
The relationship between aspect coverage and query difficulty
Validating the relationship between aspect coverage and query difficulty
Summary
Applications of Query Difficulty Estimation
Feedback: To the user and to the system
Federation and metasearch
Content enhancement using missing content analysis
Selective query expansion
Selective expansion based on query drift estimation
Adaptive use of pseudo relevance feedback
Other uses of query difficulty prediction
Summary
Summary and Conclusions
Summary
What next?
Concluding remarks
Bibliography
Authors' Biographies