FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4284707204> ?p ?o ?g. }

• W4284707204 abstract "Event argument extraction (EAE) is an important information extraction task, which aims to identify the arguments of an event described in a given text and classify the roles played by them. A key characteristic in realistic EAE data is that the instance numbers of different roles follow an obvious long-tail distribution. However, the training and evaluation paradigms of existing EAE models either prone to neglect the performance on tail roles'', or change the role instance distribution for model training to an unrealistic uniform distribution. Though some generic methods can alleviate the class imbalance in long-tail datasets, they usually sacrifice the performance of head classes'' as a trade-off. To address the above issues, we propose to train our model on realistic long-tail EAE datasets, and evaluate the average performance over all roles. Inspired by the Mixture of Experts (MOE), we propose a Routing-Balanced Dual Expert Framework (RBDEF), which divides all roles into head and tail two scopes and assigns the classifications of head and tail roles to two separate experts. In inference, each encoded instance will be allocated to one of the two experts by a routing mechanism. To reduce routing errors caused by the imbalance of role instances, we design a Balanced Routing Mechanism (BRM), which transfers several head roles to the tail expert to balance the load of routing, and employs a tri-filter routing strategy to reduce the misallocation of the tail expert's instances. To enable an effective learning of tail roles with scarce instances, we devise Target-Specialized Meta Learning (TSML) to train the tail expert. Different from other meta learning algorithms that only search a generic parameter initialization equally applying to infinite tasks, TSML can adaptively adjust its search path to obtain a specialized initialization for the tail expert, thereby expanding the benefits to the learning of tail roles. In experiments, RBDEF significantly outperforms the state-of-the-art EAE models and advanced methods for long-tail data." @default.
• W4284707204 created "2022-07-08" @default.
• W4284707204 creator A5036865453 @default.
• W4284707204 creator A5059067384 @default.
• W4284707204 creator A5060417049 @default.
• W4284707204 creator A5081675173 @default.
• W4284707204 date "2022-07-06" @default.
• W4284707204 modified "2023-09-23" @default.
• W4284707204 title "A Dual-Expert Framework for Event Argument Extraction" @default.
• W4284707204 cites W2064675550 @default.
• W4284707204 cites W2150884987 @default.
• W4284707204 cites W2250999640 @default.
• W4284707204 cites W2604941932 @default.
• W4284707204 cites W2738334887 @default.
• W4284707204 cites W2888541716 @default.
• W4284707204 cites W2890267145 @default.
• W4284707204 cites W2952437275 @default.
• W4284707204 cites W2963473975 @default.
• W4284707204 cites W2963866663 @default.
• W4284707204 cites W2983637478 @default.
• W4284707204 cites W2984323660 @default.
• W4284707204 cites W3005335434 @default.
• W4284707204 cites W3012741015 @default.
• W4284707204 cites W3034328552 @default.
• W4284707204 cites W3034709327 @default.
• W4284707204 cites W3034744380 @default.
• W4284707204 cites W3102925419 @default.
• W4284707204 cites W3103733301 @default.
• W4284707204 cites W3104597568 @default.
• W4284707204 cites W3174945605 @default.
• W4284707204 cites W3176033362 @default.
• W4284707204 cites W3188934550 @default.
• W4284707204 cites W3199464215 @default.
• W4284707204 cites W3201658153 @default.
• W4284707204 cites W3208059549 @default.
• W4284707204 cites W3212660220 @default.
• W4284707204 cites W3214600742 @default.
• W4284707204 cites W4205312537 @default.
• W4284707204 doi "https://doi.org/10.1145/3477495.3531923" @default.
• W4284707204 hasPublicationYear "2022" @default.
• W4284707204 type Work @default.
• W4284707204 citedByCount "1" @default.
• W4284707204 countsByYear W42847072042023 @default.
• W4284707204 crossrefType "proceedings-article" @default.
• W4284707204 hasAuthorship W4284707204A5036865453 @default.
• W4284707204 hasAuthorship W4284707204A5059067384 @default.
• W4284707204 hasAuthorship W4284707204A5060417049 @default.
• W4284707204 hasAuthorship W4284707204A5081675173 @default.
• W4284707204 hasConcept C119857082 @default.
• W4284707204 hasConcept C121332964 @default.
• W4284707204 hasConcept C124101348 @default.
• W4284707204 hasConcept C124952713 @default.
• W4284707204 hasConcept C142362112 @default.
• W4284707204 hasConcept C154945302 @default.
• W4284707204 hasConcept C162324750 @default.
• W4284707204 hasConcept C185592680 @default.
• W4284707204 hasConcept C187736073 @default.
• W4284707204 hasConcept C2776214188 @default.
• W4284707204 hasConcept C2777212361 @default.
• W4284707204 hasConcept C2779662365 @default.
• W4284707204 hasConcept C2780451532 @default.
• W4284707204 hasConcept C2780980858 @default.
• W4284707204 hasConcept C31258907 @default.
• W4284707204 hasConcept C41008148 @default.
• W4284707204 hasConcept C55493867 @default.
• W4284707204 hasConcept C62520636 @default.
• W4284707204 hasConcept C74172769 @default.
• W4284707204 hasConcept C98184364 @default.
• W4284707204 hasConceptScore W4284707204C119857082 @default.
• W4284707204 hasConceptScore W4284707204C121332964 @default.
• W4284707204 hasConceptScore W4284707204C124101348 @default.
• W4284707204 hasConceptScore W4284707204C124952713 @default.
• W4284707204 hasConceptScore W4284707204C142362112 @default.
• W4284707204 hasConceptScore W4284707204C154945302 @default.
• W4284707204 hasConceptScore W4284707204C162324750 @default.
• W4284707204 hasConceptScore W4284707204C185592680 @default.
• W4284707204 hasConceptScore W4284707204C187736073 @default.
• W4284707204 hasConceptScore W4284707204C2776214188 @default.
• W4284707204 hasConceptScore W4284707204C2777212361 @default.
• W4284707204 hasConceptScore W4284707204C2779662365 @default.
• W4284707204 hasConceptScore W4284707204C2780451532 @default.
• W4284707204 hasConceptScore W4284707204C2780980858 @default.
• W4284707204 hasConceptScore W4284707204C31258907 @default.
• W4284707204 hasConceptScore W4284707204C41008148 @default.
• W4284707204 hasConceptScore W4284707204C55493867 @default.
• W4284707204 hasConceptScore W4284707204C62520636 @default.
• W4284707204 hasConceptScore W4284707204C74172769 @default.
• W4284707204 hasConceptScore W4284707204C98184364 @default.
• W4284707204 hasFunder F4320321001 @default.
• W4284707204 hasLocation W42847072041 @default.
• W4284707204 hasOpenAccess W4284707204 @default.
• W4284707204 hasPrimaryLocation W42847072041 @default.
• W4284707204 hasRelatedWork W2081647779 @default.
• W4284707204 hasRelatedWork W2961085424 @default.
• W4284707204 hasRelatedWork W3032007989 @default.
• W4284707204 hasRelatedWork W3046775127 @default.
• W4284707204 hasRelatedWork W4285260836 @default.
• W4284707204 hasRelatedWork W4286629047 @default.
• W4284707204 hasRelatedWork W4306321456 @default.
• W4284707204 hasRelatedWork W4306674287 @default.

• next