TILDE: Term independent likelihood moDEl for passage re-ranking

S Zhuang, G Zuccon - Proceedings of the 44th International ACM SIGIR …, 2021 - dl.acm.org
Deep language models (deep LMs) are increasingly being used for full text retrieval or
within cascade retrieval pipelines as later-stage re-rankers. A problem with using deep LMs …

Deep query likelihood model for information retrieval

S Zhuang, H Li, G Zuccon - … Retrieval: 43rd European Conference on IR …, 2021 - Springer
The query likelihood model (QLM) for information retrieval has been thoroughly investigated
and utilised. At the basis of this method is the representation of queries and documents as …

Bootstrapped nDCG Estimation in the Presence of Unjudged Documents

M Fröbe, L Gienapp, M Potthast, M Hagen - European Conference on …, 2023 - Springer
Retrieval studies often reuse TREC collections after the corresponding tracks have passed.
Yet, a fair evaluation of new systems that retrieve documents outside the original judgment …

When measurement misleads: The limits of batch assessment of retrieval systems

J Zobel - ACM SIGIR Forum, 2023 - dl.acm.org
The discipline of information retrieval (IR) has a long history of examination of how best to
measure performance. In particular, there is an extensive literature on the practice of …

Analyzing Adversarial Attacks on Sequence-to-Sequence Relevance Models

A Parry, M Fröbe, S MacAvaney, M Potthast… - … on Information Retrieval, 2024 - Springer
Modern sequence-to-sequence relevance models like monoT5 can effectively capture
complex textual interactions between queries and documents through cross-encoding …

The Impact of Judgment Variability on the Consistency of Offline Effectiveness Measures

L Rashidi, J Zobel, A Moffat - ACM Transactions on Information Systems, 2023 - dl.acm.org
Measurement of the effectiveness of search engines is often based on use of relevance
judgments. It is well known that judgments can be inconsistent between judges, leading to …

SCAI-QReCC shared task on conversational question answering

S Vakulenko, J Kiesel, M Fröbe - arxiv preprint arxiv:2201.11094, 2022 - arxiv.org
Search-Oriented Conversational AI (SCAI) is an established venue that regularly puts a
spotlight upon the recent work advancing the field of conversational search. SCAI'21 was …

[PDF][PDF] Team openwebsearch at CLEF 2024: QuantumCLEF

M Frobe, D Alexander, GAW Hendriksen, F Schlatt… - 2024 - repository.ubn.ru.nl
We describe the OpenWebSearch group's participation in the CLEF 2024 QuantumClef IR
Feature Selection track. Our submitted runs focus on the observation that the importance of …

How Train–Test Leakage Affects Zero-Shot Retrieval

M Fröbe, C Akiki, M Potthast, M Hagen - International Symposium on …, 2022 - Springer
Neural retrieval models are often trained on (subsets of) the millions of queries of the MS
MARCO/ORCAS datasets and then tested on the 250 Robust04 queries or other TREC …

Evaluating the predictivity of IR experiments

L Rashidi, J Zobel, A Moffat - Proceedings of the 44th International ACM …, 2021 - dl.acm.org
Experimental evaluation is regarded as a critical element of any research activity in
Information Retrieval, and is typically used to support assertions of the form" Technique A …