{"id":"https://openalex.org/W7140089661","doi":"https://doi.org/10.18653/v1/2026.eacl-long.192","title":"Rethinking Reading Order: Toward Generalizable Document Understanding with LLM-based Relation Modeling","display_name":"Rethinking Reading Order: Toward Generalizable Document Understanding with LLM-based Relation Modeling","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7140089661","doi":"https://doi.org/10.18653/v1/2026.eacl-long.192"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.eacl-long.192","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.eacl-long.192","pdf_url":"https://aclanthology.org/2026.eacl-long.192.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.eacl-long.192.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5130355702","display_name":"Weishi Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Weishi Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5017709678","display_name":"Hengchang Hu","orcid":"https://orcid.org/0000-0001-7847-0641"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hengchang Hu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5002373633","display_name":"Daniel Dahlmeier","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Daniel Dahlmeier","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.39375127,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"4110","last_page":"4130"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T13629","display_name":"Text Readability and Simplification","score":0.4214000105857849,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T13629","display_name":"Text Readability and Simplification","score":0.4214000105857849,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.23090000450611115,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11550","display_name":"Text and Document Classification Technologies","score":0.0649000033736229,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/relation","display_name":"Relation (database)","score":0.6504999995231628},{"id":"https://openalex.org/keywords/reading","display_name":"Reading (process)","score":0.5358999967575073},{"id":"https://openalex.org/keywords/action","display_name":"Action (physics)","score":0.33149999380111694},{"id":"https://openalex.org/keywords/component","display_name":"Component (thermodynamics)","score":0.3149999976158142},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.2736999988555908}],"concepts":[{"id":"https://openalex.org/C25343380","wikidata":"https://www.wikidata.org/wiki/Q277521","display_name":"Relation (database)","level":2,"score":0.6504999995231628},{"id":"https://openalex.org/C554936623","wikidata":"https://www.wikidata.org/wiki/Q199657","display_name":"Reading (process)","level":2,"score":0.5358999967575073},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4887000024318695},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.41179999709129333},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.33149999380111694},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3260999917984009},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.3149999976158142},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3122999966144562},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.31150001287460327},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.2784999907016754},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.2736999988555908},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.26579999923706055},{"id":"https://openalex.org/C26760741","wikidata":"https://www.wikidata.org/wiki/Q160402","display_name":"Perception","level":2,"score":0.2549000084400177}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.eacl-long.192","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.eacl-long.192","pdf_url":"https://aclanthology.org/2026.eacl-long.192.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.eacl-long.192","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.eacl-long.192","pdf_url":"https://aclanthology.org/2026.eacl-long.192.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","score":0.8507388830184937,"display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7140089661.pdf","grobid_xml":"https://content.openalex.org/works/W7140089661.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Document":[0,96],"understanding":[1],"requires":[2],"modeling":[3],"both":[4,135],"structural":[5,78,116],"and":[6,47,67,95,118,124,129,137,146],"semantic":[7],"relationships":[8],"between":[9],"the":[10,14,50,127],"layout":[11,69],"elements":[12],"within":[13],"document,":[15],"with":[16,159],"human-perceived":[17],"reading":[18],"order":[19],"(RO)":[20],"playing":[21],"a":[22,54],"crucial":[23],"yet":[24],"often":[25],"neglected":[26],"role":[27],"compared":[28],"to":[29,63],"heuristic":[30],"OCR":[31],"sequences":[32],"used":[33],"by":[34],"most":[35],"existing":[36],"models.Previous":[37],"approaches":[38],"depend":[39],"on":[40,89],"costly,":[41],"inconsistent":[42],"human":[43,72],"annotations,":[44],"limiting":[45],"scalability":[46,128],"generalization.To":[48],"bridge":[49],"gap,":[51],"we":[52],"propose":[53],"cost-effective":[55],"paradigm":[56],"that":[57],"leverages":[58],"large":[59],"language":[60],"models":[61,140],"(LLMs)":[62],"infer":[64],"global":[65,115],"RO":[66,76],"inter-element":[68],"relations":[70],"without":[71],"supervision.By":[73],"explicitly":[74],"incorporating":[75],"as":[77],"guidance,":[79],"our":[80],"method":[81],"captures":[82],"hierarchical,":[83],"document-level":[84],"dependencies":[85],"beyond":[86],"local":[87],"adjacency.Experiments":[88],"Semantic":[90],"Entity":[91,93],"Recognition,":[92],"Linking,":[94],"Question":[97],"Answering":[98],"show":[99],"consistent":[100],"improvements":[101],"over":[102],"baseline":[103],"methods.Notably,":[104],"LLM-inferred":[105],"RO,":[106],"even":[107],"when":[108],"differing":[109],"from":[110],"ground-truth":[111],"adjacency,":[112],"provides":[113],"richer":[114],"priors":[117],"yields":[119],"superior":[120],"downstream":[121],"performance.These":[122],"results":[123],"findings":[125],"demonstrate":[126],"significance":[130],"of":[131],"RO-aware":[132],"modeling,":[133],"advancing":[134],"LLMs":[136],"lightweight":[138],"layout-aware":[139],"for":[141],"robust":[142],"document":[143],"understanding.Code,":[144],"data,":[145],"more":[147],"details":[148],"will":[149],"be":[150],"made":[151],"publicly":[152],"available":[153],"after":[154],"corporate":[155,161],"review,":[156],"in":[157],"accordance":[158],"SAP's":[160],"open-source":[162],"policy.":[163]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-03-24T00:00:00"}