Matches in SemOpenAlex for { <https://semopenalex.org/work/W4384890816> ?p ?o ?g. }
Showing items 1 to 100 of
100
with 100 items per page.
- W4384890816 abstract "Pretrained language models such as BERT have been shown to be exceptionally effective for text ranking. However, there are limited studies on how to leverage more powerful sequence-to-sequence models such as T5. Existing attempts usually formulate text ranking as a classification problem and rely on postprocessing to obtain a ranked list. In this paper, we propose RankT5 and study two T5-based ranking model structures, an encoder-decoder and an encoder-only one, so that they not only can directly output ranking scores for each query-document pair, but also can be fine-tuned with pairwise or listwise ranking losses to optimize ranking performance. Our experiments show that the proposed models with ranking losses can achieve substantial ranking performance gains on different public text ranking data sets. Moreover, ranking models fine-tuned with listwise ranking losses have better zero-shot ranking performance on out-of-domain data than models fine-tuned with classification losses." @default.
- W4384890816 created "2023-07-21" @default.
- W4384890816 creator A5011279860 @default.
- W4384890816 creator A5020968563 @default.
- W4384890816 creator A5023753874 @default.
- W4384890816 creator A5032248436 @default.
- W4384890816 creator A5042795557 @default.
- W4384890816 creator A5045420836 @default.
- W4384890816 creator A5064608039 @default.
- W4384890816 creator A5077817759 @default.
- W4384890816 creator A5079952984 @default.
- W4384890816 date "2023-07-18" @default.
- W4384890816 modified "2023-09-26" @default.
- W4384890816 title "RankT5: Fine-Tuning T5 for Text Ranking with Ranking Losses" @default.
- W4384890816 cites W119481552 @default.
- W4384890816 cites W2069870183 @default.
- W4384890816 cites W2108862644 @default.
- W4384890816 cites W2143331230 @default.
- W4384890816 cites W2912924812 @default.
- W4384890816 cites W2959353218 @default.
- W4384890816 cites W3021397474 @default.
- W4384890816 cites W3099700870 @default.
- W4384890816 cites W3100107515 @default.
- W4384890816 cites W3105817677 @default.
- W4384890816 cites W3105949871 @default.
- W4384890816 cites W3147292006 @default.
- W4384890816 cites W3153634982 @default.
- W4384890816 cites W3154498239 @default.
- W4384890816 cites W3155114168 @default.
- W4384890816 cites W3156636935 @default.
- W4384890816 cites W3194782062 @default.
- W4384890816 cites W3206455169 @default.
- W4384890816 cites W3214554611 @default.
- W4384890816 cites W4284685307 @default.
- W4384890816 cites W4285253555 @default.
- W4384890816 cites W4291127115 @default.
- W4384890816 cites W4292433389 @default.
- W4384890816 cites W4385572770 @default.
- W4384890816 cites W4385573057 @default.
- W4384890816 cites W4385573970 @default.
- W4384890816 doi "https://doi.org/10.1145/3539618.3592047" @default.
- W4384890816 hasPublicationYear "2023" @default.
- W4384890816 type Work @default.
- W4384890816 citedByCount "0" @default.
- W4384890816 crossrefType "proceedings-article" @default.
- W4384890816 hasAuthorship W4384890816A5011279860 @default.
- W4384890816 hasAuthorship W4384890816A5020968563 @default.
- W4384890816 hasAuthorship W4384890816A5023753874 @default.
- W4384890816 hasAuthorship W4384890816A5032248436 @default.
- W4384890816 hasAuthorship W4384890816A5042795557 @default.
- W4384890816 hasAuthorship W4384890816A5045420836 @default.
- W4384890816 hasAuthorship W4384890816A5064608039 @default.
- W4384890816 hasAuthorship W4384890816A5077817759 @default.
- W4384890816 hasAuthorship W4384890816A5079952984 @default.
- W4384890816 hasBestOaLocation W43848908161 @default.
- W4384890816 hasConcept C111919701 @default.
- W4384890816 hasConcept C118505674 @default.
- W4384890816 hasConcept C119857082 @default.
- W4384890816 hasConcept C124101348 @default.
- W4384890816 hasConcept C124975894 @default.
- W4384890816 hasConcept C153083717 @default.
- W4384890816 hasConcept C154945302 @default.
- W4384890816 hasConcept C184898388 @default.
- W4384890816 hasConcept C189430467 @default.
- W4384890816 hasConcept C23123220 @default.
- W4384890816 hasConcept C2778112365 @default.
- W4384890816 hasConcept C41008148 @default.
- W4384890816 hasConcept C54355233 @default.
- W4384890816 hasConcept C86803240 @default.
- W4384890816 hasConceptScore W4384890816C111919701 @default.
- W4384890816 hasConceptScore W4384890816C118505674 @default.
- W4384890816 hasConceptScore W4384890816C119857082 @default.
- W4384890816 hasConceptScore W4384890816C124101348 @default.
- W4384890816 hasConceptScore W4384890816C124975894 @default.
- W4384890816 hasConceptScore W4384890816C153083717 @default.
- W4384890816 hasConceptScore W4384890816C154945302 @default.
- W4384890816 hasConceptScore W4384890816C184898388 @default.
- W4384890816 hasConceptScore W4384890816C189430467 @default.
- W4384890816 hasConceptScore W4384890816C23123220 @default.
- W4384890816 hasConceptScore W4384890816C2778112365 @default.
- W4384890816 hasConceptScore W4384890816C41008148 @default.
- W4384890816 hasConceptScore W4384890816C54355233 @default.
- W4384890816 hasConceptScore W4384890816C86803240 @default.
- W4384890816 hasLocation W43848908161 @default.
- W4384890816 hasLocation W43848908162 @default.
- W4384890816 hasOpenAccess W4384890816 @default.
- W4384890816 hasPrimaryLocation W43848908161 @default.
- W4384890816 hasRelatedWork W1546661187 @default.
- W4384890816 hasRelatedWork W2021630848 @default.
- W4384890816 hasRelatedWork W2031812225 @default.
- W4384890816 hasRelatedWork W2069238515 @default.
- W4384890816 hasRelatedWork W2100679666 @default.
- W4384890816 hasRelatedWork W2105089228 @default.
- W4384890816 hasRelatedWork W3017018726 @default.
- W4384890816 hasRelatedWork W3194506168 @default.
- W4384890816 hasRelatedWork W3211265021 @default.
- W4384890816 hasRelatedWork W71818576 @default.
- W4384890816 isParatext "false" @default.
- W4384890816 isRetracted "false" @default.
- W4384890816 workType "article" @default.