Matches in SemOpenAlex for { <https://semopenalex.org/work/W3136510012> ?p ?o ?g. }
Showing items 1 to 80 of
80
with 100 items per page.
- W3136510012 abstract "This paper proposes two novel approaches to supplement BERT for making negation-word (NW) entailment judgment on a Chinese Social Studies QA dataset. Recently, BERT has shown excellent performance (and even outperforms human) on several natural language inference tasks. However, BERT is found to achieve the remarkable results mainly via utilizing the surface features (such as lexical distribution bias) that are unnoticed by humans. In our test, BERT’s performance would degrade significantly if it is tested on a NW involved dataset in which the influence of lexical distribution bias has been removed. Since a single unmatched NW could toggle the overall judgment, we propose a Negation-Word-Aggregative-Pattern (NWAP) for reflecting the NW matching status, and use it to supplement BERT. Our first approach then supplements BERT with a NW toggling module, which decides if BERT’s answer should be toggled according to the NWAP. The second approach concatenates the embedding of above NWAP with BERT’s output vectors, and then feed them into a feedforward (FF) neural network (NN) to make the final judgment. Experiments show our toggling module outperforms BERT by 59% accuracy on those lexical-bias removed NW data-set. With adversary training data added to both BERT and our FF NN, our model still outperforms BERT by 23% accuracy." @default.
- W3136510012 created "2021-03-29" @default.
- W3136510012 creator A5070399630 @default.
- W3136510012 creator A5088900247 @default.
- W3136510012 date "2020-12-01" @default.
- W3136510012 modified "2023-09-23" @default.
- W3136510012 title "Making Negation-word Entailment Judgment via Supplementing BERT with Aggregative Pattern" @default.
- W3136510012 cites W1840435438 @default.
- W3136510012 cites W1961993270 @default.
- W3136510012 cites W2080278171 @default.
- W3136510012 cites W2130158090 @default.
- W3136510012 cites W2413794162 @default.
- W3136510012 cites W2415204069 @default.
- W3136510012 cites W2891012317 @default.
- W3136510012 cites W2942354996 @default.
- W3136510012 cites W2962843521 @default.
- W3136510012 cites W2963341956 @default.
- W3136510012 cites W2963719234 @default.
- W3136510012 cites W2965373594 @default.
- W3136510012 cites W2970438301 @default.
- W3136510012 cites W3034584102 @default.
- W3136510012 cites W2525127255 @default.
- W3136510012 cites W2971885209 @default.
- W3136510012 doi "https://doi.org/10.1109/taai51410.2020.00012" @default.
- W3136510012 hasPublicationYear "2020" @default.
- W3136510012 type Work @default.
- W3136510012 sameAs 3136510012 @default.
- W3136510012 citedByCount "0" @default.
- W3136510012 crossrefType "proceedings-article" @default.
- W3136510012 hasAuthorship W3136510012A5070399630 @default.
- W3136510012 hasAuthorship W3136510012A5088900247 @default.
- W3136510012 hasConcept C105795698 @default.
- W3136510012 hasConcept C134752490 @default.
- W3136510012 hasConcept C138885662 @default.
- W3136510012 hasConcept C154945302 @default.
- W3136510012 hasConcept C165064840 @default.
- W3136510012 hasConcept C199360897 @default.
- W3136510012 hasConcept C204321447 @default.
- W3136510012 hasConcept C2185349 @default.
- W3136510012 hasConcept C2776214188 @default.
- W3136510012 hasConcept C33923547 @default.
- W3136510012 hasConcept C41008148 @default.
- W3136510012 hasConcept C41608201 @default.
- W3136510012 hasConcept C41895202 @default.
- W3136510012 hasConcept C50644808 @default.
- W3136510012 hasConcept C90805587 @default.
- W3136510012 hasConcept C95318506 @default.
- W3136510012 hasConceptScore W3136510012C105795698 @default.
- W3136510012 hasConceptScore W3136510012C134752490 @default.
- W3136510012 hasConceptScore W3136510012C138885662 @default.
- W3136510012 hasConceptScore W3136510012C154945302 @default.
- W3136510012 hasConceptScore W3136510012C165064840 @default.
- W3136510012 hasConceptScore W3136510012C199360897 @default.
- W3136510012 hasConceptScore W3136510012C204321447 @default.
- W3136510012 hasConceptScore W3136510012C2185349 @default.
- W3136510012 hasConceptScore W3136510012C2776214188 @default.
- W3136510012 hasConceptScore W3136510012C33923547 @default.
- W3136510012 hasConceptScore W3136510012C41008148 @default.
- W3136510012 hasConceptScore W3136510012C41608201 @default.
- W3136510012 hasConceptScore W3136510012C41895202 @default.
- W3136510012 hasConceptScore W3136510012C50644808 @default.
- W3136510012 hasConceptScore W3136510012C90805587 @default.
- W3136510012 hasConceptScore W3136510012C95318506 @default.
- W3136510012 hasLocation W31365100121 @default.
- W3136510012 hasOpenAccess W3136510012 @default.
- W3136510012 hasPrimaryLocation W31365100121 @default.
- W3136510012 hasRelatedWork W10132401 @default.
- W3136510012 hasRelatedWork W11133913 @default.
- W3136510012 hasRelatedWork W12091112 @default.
- W3136510012 hasRelatedWork W12963412 @default.
- W3136510012 hasRelatedWork W13287333 @default.
- W3136510012 hasRelatedWork W2061806 @default.
- W3136510012 hasRelatedWork W706006 @default.
- W3136510012 hasRelatedWork W7849509 @default.
- W3136510012 hasRelatedWork W90294 @default.
- W3136510012 hasRelatedWork W988378 @default.
- W3136510012 isParatext "false" @default.
- W3136510012 isRetracted "false" @default.
- W3136510012 magId "3136510012" @default.
- W3136510012 workType "article" @default.