Matches in SemOpenAlex for { <https://semopenalex.org/work/W2889581211> ?p ?o ?g. }
- W2889581211 abstract "Learning a matching function between two text sequences is a long standing problem in NLP research. This task enables many potential applications such as question answering and paraphrase identification. This paper proposes Co-Stack Residual Affinity Networks (CSRAN), a new and universal neural architecture for this problem. CSRAN is a deep architecture, involving stacked (multi-layered) recurrent encoders. Stacked/Deep architectures are traditionally difficult to train, due to the inherent weaknesses such as difficulty with feature propagation and vanishing gradients. CSRAN incorporates two novel components to take advantage of the stacked architecture. Firstly, it introduces a new bidirectional alignment mechanism that learns affinity weights by fusing sequence pairs across stacked hierarchies. Secondly, it leverages a multi-level attention refinement component between stacked recurrent layers. The key intuition is that, by leveraging information across all network hierarchies, we can not only improve gradient flow but also improve overall performance. We conduct extensive experiments on six well-studied text sequence matching datasets, achieving state-of-the-art performance on all." @default.
- W2889581211 created "2018-09-27" @default.
- W2889581211 creator A5026059624 @default.
- W2889581211 creator A5050386762 @default.
- W2889581211 creator A5057786383 @default.
- W2889581211 date "2018-01-01" @default.
- W2889581211 modified "2023-10-16" @default.
- W2889581211 title "Co-Stack Residual Affinity Networks with Multi-level Attention Refinement for Matching Text Sequences" @default.
- W2889581211 cites W1533861849 @default.
- W2889581211 cites W1591825359 @default.
- W2889581211 cites W1840435438 @default.
- W2889581211 cites W1966443646 @default.
- W2889581211 cites W2118463056 @default.
- W2889581211 cites W2120735855 @default.
- W2889581211 cites W2130942839 @default.
- W2889581211 cites W2131876387 @default.
- W2889581211 cites W2143612262 @default.
- W2889581211 cites W2153702313 @default.
- W2889581211 cites W2194775991 @default.
- W2889581211 cites W2250539671 @default.
- W2889581211 cites W2251427843 @default.
- W2889581211 cites W2280395961 @default.
- W2889581211 cites W2295739661 @default.
- W2889581211 cites W2338325072 @default.
- W2889581211 cites W2413794162 @default.
- W2889581211 cites W2469060249 @default.
- W2889581211 cites W2511929605 @default.
- W2889581211 cites W2516930406 @default.
- W2889581211 cites W2551396370 @default.
- W2889581211 cites W2573379274 @default.
- W2889581211 cites W2608787653 @default.
- W2889581211 cites W2612867916 @default.
- W2889581211 cites W2749909881 @default.
- W2889581211 cites W2760753016 @default.
- W2889581211 cites W2767501021 @default.
- W2889581211 cites W2767857566 @default.
- W2889581211 cites W2782363479 @default.
- W2889581211 cites W2788496822 @default.
- W2889581211 cites W2808281579 @default.
- W2889581211 cites W2809057686 @default.
- W2889581211 cites W2951528484 @default.
- W2889581211 cites W2952113915 @default.
- W2889581211 cites W2962685628 @default.
- W2889581211 cites W2962739339 @default.
- W2889581211 cites W2962854379 @default.
- W2889581211 cites W2963053846 @default.
- W2889581211 cites W2963077723 @default.
- W2889581211 cites W2963446712 @default.
- W2889581211 cites W2963508788 @default.
- W2889581211 cites W2963615308 @default.
- W2889581211 cites W2963719234 @default.
- W2889581211 cites W2963756346 @default.
- W2889581211 cites W2963871484 @default.
- W2889581211 cites W2964012472 @default.
- W2889581211 cites W2964026924 @default.
- W2889581211 cites W2964084166 @default.
- W2889581211 cites W2964121744 @default.
- W2889581211 cites W2964308564 @default.
- W2889581211 cites W3101747393 @default.
- W2889581211 doi "https://doi.org/10.18653/v1/d18-1479" @default.
- W2889581211 hasPublicationYear "2018" @default.
- W2889581211 type Work @default.
- W2889581211 sameAs 2889581211 @default.
- W2889581211 citedByCount "20" @default.
- W2889581211 countsByYear W28895812112018 @default.
- W2889581211 countsByYear W28895812112019 @default.
- W2889581211 countsByYear W28895812112020 @default.
- W2889581211 countsByYear W28895812112021 @default.
- W2889581211 countsByYear W28895812112022 @default.
- W2889581211 countsByYear W28895812112023 @default.
- W2889581211 crossrefType "proceedings-article" @default.
- W2889581211 hasAuthorship W2889581211A5026059624 @default.
- W2889581211 hasAuthorship W2889581211A5050386762 @default.
- W2889581211 hasAuthorship W2889581211A5057786383 @default.
- W2889581211 hasBestOaLocation W28895812111 @default.
- W2889581211 hasConcept C105795698 @default.
- W2889581211 hasConcept C111472728 @default.
- W2889581211 hasConcept C111919701 @default.
- W2889581211 hasConcept C11413529 @default.
- W2889581211 hasConcept C118505674 @default.
- W2889581211 hasConcept C132010649 @default.
- W2889581211 hasConcept C138885662 @default.
- W2889581211 hasConcept C154945302 @default.
- W2889581211 hasConcept C155512373 @default.
- W2889581211 hasConcept C165064840 @default.
- W2889581211 hasConcept C33923547 @default.
- W2889581211 hasConcept C41008148 @default.
- W2889581211 hasConceptScore W2889581211C105795698 @default.
- W2889581211 hasConceptScore W2889581211C111472728 @default.
- W2889581211 hasConceptScore W2889581211C111919701 @default.
- W2889581211 hasConceptScore W2889581211C11413529 @default.
- W2889581211 hasConceptScore W2889581211C118505674 @default.
- W2889581211 hasConceptScore W2889581211C132010649 @default.
- W2889581211 hasConceptScore W2889581211C138885662 @default.
- W2889581211 hasConceptScore W2889581211C154945302 @default.
- W2889581211 hasConceptScore W2889581211C155512373 @default.
- W2889581211 hasConceptScore W2889581211C165064840 @default.
- W2889581211 hasConceptScore W2889581211C33923547 @default.
- W2889581211 hasConceptScore W2889581211C41008148 @default.
- W2889581211 hasLocation W28895812111 @default.