English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

This paper reports on an effort of reproducing the organizers' baseline aswell as the top performing participant submission at the 2021 edition of theTREC Conversational Assistance track. TREC systems are commonly regarded asreference points for effectiveness comparison. Yet, the papers accompanyingthem have less strict requirements than peer-reviewed publications, which canmake reproducibility challenging. Our results indicate that key practicalinformation is indeed missing. While the results can be reproduced within a 19%relative margin with respect to the main evaluation measure, the relativedifference between the baseline and the top performing approach shrinks fromthe reported 18% to 5%. Additionally, we report on a new set of experimentsaimed at understanding the impact of various pipeline components. We show thatend-to-end system performance can indeed benefit from advanced retrievaltechniques in either stage of a two-stage retrieval pipeline. We also measurethe impact of the dataset used for fine-tuning the query rewriter and find thatemploying different query rewriting methods in different stages of theretrieval pipeline might be beneficial. Moreover, these results are shown togeneralize across the 2020 and 2021 editions of the track. We conclude ourstudy with a list of lessons learned and practical suggestions.

From Baseline to Top Performer: A Reproducibility Study of Approaches at the TREC 2021 Conversational Assistance Track