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/

Global: Zendy Plus annual

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

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Search engine ranking pipelines are commonly based on large ensembles ofmachine-learned decision trees. The tight constraints on query response timerecently motivated researchers to investigate algorithms to make faster thetraversal of the additive ensemble or to early terminate the evaluation ofdocuments that are unlikely to be ranked among the top-k. In this paper, weinvestigate the novel problem of \textit{query-level early exiting}, aimed atdeciding the profitability of early stopping the traversal of the rankingensemble for all the candidate documents to be scored for a query, by simplyreturning a ranking based on the additive scores computed by a limited portionof the ensemble. Besides the obvious advantage on query latency and throughput,we address the possible positive impact of query-level early exiting on rankingeffectiveness. To this end, we study the actual contribution of incrementalportions of the tree ensemble to the ranking of the top-k documents scored fora given query. Our main finding is that queries exhibit different behaviors asscores are accumulated during the traversal of the ensemble and thatquery-level early stopping can remarkably improve ranking quality. We present areproducible and comprehensive experimental evaluation, conducted on two publicdatasets, showing that query-level early exiting achieves an overall gain of upto 7.5% in terms of NDCG@10 with a speedup of the scoring process of up to2.2x.

Query-level Early Exit for Additive Learning-to-Rank Ensembles