Publications

Query Performance Prediction through Retrieval Coherency

Negar Arabzadeh and Amin Bigdeli and Morteza Zihayat and Ebrahim Bagheri
Reference:
Negar Arabzadeh; Amin Bigdeli; Morteza Zihayat and Ebrahim Bagheri Query Performance Prediction through Retrieval Coherency. In 43rd European Conference on IR Research (ECIR 2021), 2021.
Links to Publication:
Abstract:
Post-retrieval Query Performance Prediction (QPP) methods benefit from the characteristics of the retrieved set of documents to determine query difficulty. While existing works have investigated the relation between query and retrieved document spaces, as well as retrieved document scores, the association between the retrieved documents themselves, referred to as coherency, has not been extensively investigated for QPP. We propose that the coherence of the retrieved documents can be formalized as a function of the characteristics of a network that represents the associations between these documents. We show that existing coherence metrics are a special case of our formalization of coherence. Based on experiments on three corpora, namely Robust04, Gov2 and ClueWeb09 and their TREC topics, we show that our coherence measures outperform existing metrics in the literature and are able to significantly improve the performance of state of the art QPP methods.
Bibtex Entry:
@inproceedings{ecir2021c, author = {Negar Arabzadeh and Amin Bigdeli and Morteza Zihayat and Ebrahim Bagheri}, title = {Query Performance Prediction through Retrieval Coherency}, booktitle = {43rd European Conference on IR Research (ECIR 2021)}, year = {2021}, abstract = {Post-retrieval Query Performance Prediction (QPP) methods benefit from the characteristics of the retrieved set of documents to determine query difficulty. While existing works have investigated the relation between query and retrieved document spaces, as well as retrieved document scores, the association between the retrieved documents themselves, referred to as coherency, has not been extensively investigated for QPP. We propose that the coherence of the retrieved documents can be formalized as a function of the characteristics of a network that represents the associations between these documents. We show that existing coherence metrics are a special case of our formalization of coherence. Based on experiments on three corpora, namely Robust04, Gov2 and ClueWeb09 and their TREC topics, we show that our coherence measures outperform existing metrics in the literature and are able to significantly improve the performance of state of the art QPP methods.} }




Powered by WordPress