Matches in SemOpenAlex for { <https://semopenalex.org/work/W3175378952> ?p ?o ?g. }
Showing items 1 to 94 of
94
with 100 items per page.
- W3175378952 abstract "State-of-the-art summarization systems are trained and evaluated on massive datasets scraped from the web. Despite their prevalence, we know very little about the underlying characteristics (data noise, summarization complexity, etc.) of these datasets, and how these affect system performance and the reliability of automatic metrics like ROUGE. In this study, we manually analyze 600 samples from three popular summarization datasets. Our study is driven by a six-class typology which captures different noise types (missing facts, entities) and degrees of summarization difficulty (extractive, abstractive). We follow with a thorough analysis of 27 state-of-the-art summarization models and 5 popular metrics, and report our key insights: (1) Datasets have distinct data quality and complexity distributions, which can be traced back to their collection process. (2) The performance of models and reliability of metrics is dependent on sample complexity. (3) Faithful summaries often receive low scores because of the poor diversity of references. We release the code, annotated data and model outputs." @default.
- W3175378952 created "2021-07-05" @default.
- W3175378952 creator A5041937687 @default.
- W3175378952 creator A5068329786 @default.
- W3175378952 creator A5090555712 @default.
- W3175378952 date "2021-01-01" @default.
- W3175378952 modified "2023-10-05" @default.
- W3175378952 title "How well do you know your summarization datasets?" @default.
- W3175378952 cites W1501617060 @default.
- W3175378952 cites W1544827683 @default.
- W3175378952 cites W1939882552 @default.
- W3175378952 cites W2101105183 @default.
- W3175378952 cites W2102065370 @default.
- W3175378952 cites W2154652894 @default.
- W3175378952 cites W2552839021 @default.
- W3175378952 cites W2574535369 @default.
- W3175378952 cites W2606974598 @default.
- W3175378952 cites W2768957049 @default.
- W3175378952 cites W2799149803 @default.
- W3175378952 cites W2888482885 @default.
- W3175378952 cites W2889518897 @default.
- W3175378952 cites W2890419630 @default.
- W3175378952 cites W2894293047 @default.
- W3175378952 cites W2896739098 @default.
- W3175378952 cites W2924690340 @default.
- W3175378952 cites W2950692458 @default.
- W3175378952 cites W2951265142 @default.
- W3175378952 cites W2953280096 @default.
- W3175378952 cites W2962788840 @default.
- W3175378952 cites W2962809918 @default.
- W3175378952 cites W2962849707 @default.
- W3175378952 cites W2962965405 @default.
- W3175378952 cites W2962972512 @default.
- W3175378952 cites W2963385935 @default.
- W3175378952 cites W2963721761 @default.
- W3175378952 cites W2963929190 @default.
- W3175378952 cites W2964028111 @default.
- W3175378952 cites W2970419734 @default.
- W3175378952 cites W2970785793 @default.
- W3175378952 cites W2971034336 @default.
- W3175378952 cites W2971274815 @default.
- W3175378952 cites W2985619053 @default.
- W3175378952 cites W2996403597 @default.
- W3175378952 cites W2998640556 @default.
- W3175378952 cites W3006661753 @default.
- W3175378952 cites W3016234399 @default.
- W3175378952 cites W3034353423 @default.
- W3175378952 cites W3034715004 @default.
- W3175378952 cites W3035050380 @default.
- W3175378952 cites W3099286868 @default.
- W3175378952 cites W3100439847 @default.
- W3175378952 cites W3101017384 @default.
- W3175378952 cites W3103298261 @default.
- W3175378952 cites W3106234277 @default.
- W3175378952 cites W3170083118 @default.
- W3175378952 cites W3170658419 @default.
- W3175378952 doi "https://doi.org/10.18653/v1/2021.findings-acl.303" @default.
- W3175378952 hasPublicationYear "2021" @default.
- W3175378952 type Work @default.
- W3175378952 sameAs 3175378952 @default.
- W3175378952 citedByCount "3" @default.
- W3175378952 countsByYear W31753789522022 @default.
- W3175378952 countsByYear W31753789522023 @default.
- W3175378952 crossrefType "proceedings-article" @default.
- W3175378952 hasAuthorship W3175378952A5041937687 @default.
- W3175378952 hasAuthorship W3175378952A5068329786 @default.
- W3175378952 hasAuthorship W3175378952A5090555712 @default.
- W3175378952 hasBestOaLocation W31753789521 @default.
- W3175378952 hasConcept C134714966 @default.
- W3175378952 hasConcept C170858558 @default.
- W3175378952 hasConcept C23123220 @default.
- W3175378952 hasConcept C41008148 @default.
- W3175378952 hasConceptScore W3175378952C134714966 @default.
- W3175378952 hasConceptScore W3175378952C170858558 @default.
- W3175378952 hasConceptScore W3175378952C23123220 @default.
- W3175378952 hasConceptScore W3175378952C41008148 @default.
- W3175378952 hasLocation W31753789521 @default.
- W3175378952 hasLocation W31753789522 @default.
- W3175378952 hasOpenAccess W3175378952 @default.
- W3175378952 hasPrimaryLocation W31753789521 @default.
- W3175378952 hasRelatedWork W132250100 @default.
- W3175378952 hasRelatedWork W2093597205 @default.
- W3175378952 hasRelatedWork W2104677027 @default.
- W3175378952 hasRelatedWork W2325978870 @default.
- W3175378952 hasRelatedWork W2380641910 @default.
- W3175378952 hasRelatedWork W2389846579 @default.
- W3175378952 hasRelatedWork W2392495745 @default.
- W3175378952 hasRelatedWork W2725657302 @default.
- W3175378952 hasRelatedWork W4200075185 @default.
- W3175378952 hasRelatedWork W65350537 @default.
- W3175378952 isParatext "false" @default.
- W3175378952 isRetracted "false" @default.
- W3175378952 magId "3175378952" @default.
- W3175378952 workType "article" @default.