Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386566563> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4386566563 abstract "Documents that consist of diverse templates and exhibit complex spatial structures pose a challenge for document entity classification. We propose KNN-Former, which incorporates a new kind of spatial bias in attention calculation based on the K-nearest-neighbor (KNN) graph of document entities. We limit entities’ attention only to their local radius defined by the KNN graph. We also use combinatorial matching to address the one-to-one mapping property that exists in many documents, where one field has only one corresponding entity. Moreover, our method is highly parameter-efficient compared to existing approaches in terms of the number of trainable parameters. Despite this, experiments across various datasets show our method outperforms baselines in most entity types. Many real-world documents exhibit combinatorial properties which can be leveraged as inductive biases to improve extraction accuracy, but existing datasets do not cover these documents.To facilitate future research into these types of documents, we release a new ID document dataset that covers diverse templates and languages. We also release enhanced annotations for an existing dataset." @default.
- W4386566563 created "2023-09-10" @default.
- W4386566563 creator A5027098865 @default.
- W4386566563 creator A5033376109 @default.
- W4386566563 creator A5043925137 @default.
- W4386566563 creator A5050008056 @default.
- W4386566563 creator A5052637103 @default.
- W4386566563 creator A5054230379 @default.
- W4386566563 creator A5062400907 @default.
- W4386566563 creator A5079951463 @default.
- W4386566563 date "2023-01-01" @default.
- W4386566563 modified "2023-10-16" @default.
- W4386566563 title "Lightweight Spatial Modeling for Combinatorial Information Extraction From Documents" @default.
- W4386566563 doi "https://doi.org/10.18653/v1/2023.findings-eacl.108" @default.
- W4386566563 hasPublicationYear "2023" @default.
- W4386566563 type Work @default.
- W4386566563 citedByCount "0" @default.
- W4386566563 crossrefType "proceedings-article" @default.
- W4386566563 hasAuthorship W4386566563A5027098865 @default.
- W4386566563 hasAuthorship W4386566563A5033376109 @default.
- W4386566563 hasAuthorship W4386566563A5043925137 @default.
- W4386566563 hasAuthorship W4386566563A5050008056 @default.
- W4386566563 hasAuthorship W4386566563A5052637103 @default.
- W4386566563 hasAuthorship W4386566563A5054230379 @default.
- W4386566563 hasAuthorship W4386566563A5062400907 @default.
- W4386566563 hasAuthorship W4386566563A5079951463 @default.
- W4386566563 hasBestOaLocation W43865665631 @default.
- W4386566563 hasConcept C105795698 @default.
- W4386566563 hasConcept C111472728 @default.
- W4386566563 hasConcept C113238511 @default.
- W4386566563 hasConcept C124101348 @default.
- W4386566563 hasConcept C132525143 @default.
- W4386566563 hasConcept C138885662 @default.
- W4386566563 hasConcept C154945302 @default.
- W4386566563 hasConcept C165064840 @default.
- W4386566563 hasConcept C189950617 @default.
- W4386566563 hasConcept C195807954 @default.
- W4386566563 hasConcept C199360897 @default.
- W4386566563 hasConcept C202444582 @default.
- W4386566563 hasConcept C23123220 @default.
- W4386566563 hasConcept C33923547 @default.
- W4386566563 hasConcept C41008148 @default.
- W4386566563 hasConcept C80444323 @default.
- W4386566563 hasConcept C82714645 @default.
- W4386566563 hasConcept C9652623 @default.
- W4386566563 hasConceptScore W4386566563C105795698 @default.
- W4386566563 hasConceptScore W4386566563C111472728 @default.
- W4386566563 hasConceptScore W4386566563C113238511 @default.
- W4386566563 hasConceptScore W4386566563C124101348 @default.
- W4386566563 hasConceptScore W4386566563C132525143 @default.
- W4386566563 hasConceptScore W4386566563C138885662 @default.
- W4386566563 hasConceptScore W4386566563C154945302 @default.
- W4386566563 hasConceptScore W4386566563C165064840 @default.
- W4386566563 hasConceptScore W4386566563C189950617 @default.
- W4386566563 hasConceptScore W4386566563C195807954 @default.
- W4386566563 hasConceptScore W4386566563C199360897 @default.
- W4386566563 hasConceptScore W4386566563C202444582 @default.
- W4386566563 hasConceptScore W4386566563C23123220 @default.
- W4386566563 hasConceptScore W4386566563C33923547 @default.
- W4386566563 hasConceptScore W4386566563C41008148 @default.
- W4386566563 hasConceptScore W4386566563C80444323 @default.
- W4386566563 hasConceptScore W4386566563C82714645 @default.
- W4386566563 hasConceptScore W4386566563C9652623 @default.
- W4386566563 hasLocation W43865665631 @default.
- W4386566563 hasOpenAccess W4386566563 @default.
- W4386566563 hasPrimaryLocation W43865665631 @default.
- W4386566563 hasRelatedWork W1788528807 @default.
- W4386566563 hasRelatedWork W1896367721 @default.
- W4386566563 hasRelatedWork W1997732015 @default.
- W4386566563 hasRelatedWork W2153799433 @default.
- W4386566563 hasRelatedWork W2352337653 @default.
- W4386566563 hasRelatedWork W2357241418 @default.
- W4386566563 hasRelatedWork W2367301249 @default.
- W4386566563 hasRelatedWork W2379157006 @default.
- W4386566563 hasRelatedWork W2393978999 @default.
- W4386566563 hasRelatedWork W2725657302 @default.
- W4386566563 isParatext "false" @default.
- W4386566563 isRetracted "false" @default.
- W4386566563 workType "article" @default.