Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385571760> ?p ?o ?g. }
Showing items 1 to 81 of
81
with 100 items per page.
- W4385571760 abstract "Existing metrics for evaluating the quality of automatically generated questions such as BLEU, ROUGE, BERTScore, and BLEURT compare the reference and predicted questions, providing a high score when there is a considerable lexical overlap or semantic similarity between the candidate and the reference questions. This approach has two major shortcomings. First, we need expensive human-provided reference questions. Second, it penalises valid questions that may not have high lexical or semantic similarity to the reference questions. In this paper, we propose a new metric, RQUGE, based on the answerability of the candidate question given the context. The metric consists of a question-answering and a span scorer modules, using pre-trained models from existing literature, thus it can be used without any further training. We demonstrate that RQUGE has a higher correlation with human judgment without relying on the reference question. Additionally, RQUGE is shown to be more robust to several adversarial corruptions. Furthermore, we illustrate that we can significantly improve the performance of QA models on out-of-domain datasets by fine-tuning on synthetic data generated by a question generation model and reranked by RQUGE." @default.
- W4385571760 created "2023-08-05" @default.
- W4385571760 creator A5003287416 @default.
- W4385571760 creator A5004605673 @default.
- W4385571760 creator A5067220251 @default.
- W4385571760 creator A5081443390 @default.
- W4385571760 creator A5083185771 @default.
- W4385571760 creator A5084321238 @default.
- W4385571760 creator A5089419558 @default.
- W4385571760 date "2023-01-01" @default.
- W4385571760 modified "2023-09-27" @default.
- W4385571760 title "RQUGE: Reference-Free Metric for Evaluating Question Generation by Answering the Question" @default.
- W4385571760 doi "https://doi.org/10.18653/v1/2023.findings-acl.428" @default.
- W4385571760 hasPublicationYear "2023" @default.
- W4385571760 type Work @default.
- W4385571760 citedByCount "0" @default.
- W4385571760 crossrefType "proceedings-article" @default.
- W4385571760 hasAuthorship W4385571760A5003287416 @default.
- W4385571760 hasAuthorship W4385571760A5004605673 @default.
- W4385571760 hasAuthorship W4385571760A5067220251 @default.
- W4385571760 hasAuthorship W4385571760A5081443390 @default.
- W4385571760 hasAuthorship W4385571760A5083185771 @default.
- W4385571760 hasAuthorship W4385571760A5084321238 @default.
- W4385571760 hasAuthorship W4385571760A5089419558 @default.
- W4385571760 hasBestOaLocation W43855717601 @default.
- W4385571760 hasConcept C103278499 @default.
- W4385571760 hasConcept C115961682 @default.
- W4385571760 hasConcept C119857082 @default.
- W4385571760 hasConcept C130318100 @default.
- W4385571760 hasConcept C134306372 @default.
- W4385571760 hasConcept C151730666 @default.
- W4385571760 hasConcept C154945302 @default.
- W4385571760 hasConcept C162324750 @default.
- W4385571760 hasConcept C176217482 @default.
- W4385571760 hasConcept C204321447 @default.
- W4385571760 hasConcept C21547014 @default.
- W4385571760 hasConcept C23123220 @default.
- W4385571760 hasConcept C2779343474 @default.
- W4385571760 hasConcept C2993776861 @default.
- W4385571760 hasConcept C33923547 @default.
- W4385571760 hasConcept C36503486 @default.
- W4385571760 hasConcept C37736160 @default.
- W4385571760 hasConcept C41008148 @default.
- W4385571760 hasConcept C44291984 @default.
- W4385571760 hasConcept C86803240 @default.
- W4385571760 hasConceptScore W4385571760C103278499 @default.
- W4385571760 hasConceptScore W4385571760C115961682 @default.
- W4385571760 hasConceptScore W4385571760C119857082 @default.
- W4385571760 hasConceptScore W4385571760C130318100 @default.
- W4385571760 hasConceptScore W4385571760C134306372 @default.
- W4385571760 hasConceptScore W4385571760C151730666 @default.
- W4385571760 hasConceptScore W4385571760C154945302 @default.
- W4385571760 hasConceptScore W4385571760C162324750 @default.
- W4385571760 hasConceptScore W4385571760C176217482 @default.
- W4385571760 hasConceptScore W4385571760C204321447 @default.
- W4385571760 hasConceptScore W4385571760C21547014 @default.
- W4385571760 hasConceptScore W4385571760C23123220 @default.
- W4385571760 hasConceptScore W4385571760C2779343474 @default.
- W4385571760 hasConceptScore W4385571760C2993776861 @default.
- W4385571760 hasConceptScore W4385571760C33923547 @default.
- W4385571760 hasConceptScore W4385571760C36503486 @default.
- W4385571760 hasConceptScore W4385571760C37736160 @default.
- W4385571760 hasConceptScore W4385571760C41008148 @default.
- W4385571760 hasConceptScore W4385571760C44291984 @default.
- W4385571760 hasConceptScore W4385571760C86803240 @default.
- W4385571760 hasLocation W43855717601 @default.
- W4385571760 hasOpenAccess W4385571760 @default.
- W4385571760 hasPrimaryLocation W43855717601 @default.
- W4385571760 hasRelatedWork W1592081640 @default.
- W4385571760 hasRelatedWork W207304934 @default.
- W4385571760 hasRelatedWork W2357960838 @default.
- W4385571760 hasRelatedWork W2368388617 @default.
- W4385571760 hasRelatedWork W2547211351 @default.
- W4385571760 hasRelatedWork W2559338413 @default.
- W4385571760 hasRelatedWork W2564015900 @default.
- W4385571760 hasRelatedWork W2963829519 @default.
- W4385571760 hasRelatedWork W3192083251 @default.
- W4385571760 hasRelatedWork W4308242723 @default.
- W4385571760 isParatext "false" @default.
- W4385571760 isRetracted "false" @default.
- W4385571760 workType "article" @default.