{"id":"https://openalex.org/W7166706237","doi":"https://doi.org/10.48550/arxiv.2606.29808","title":"Making Multimodal LLMs Reliable Chart Data Extractors: A Benchmark and Training Framework","display_name":"Making Multimodal LLMs Reliable Chart Data Extractors: A Benchmark and Training Framework","publication_year":2026,"publication_date":"2026-06-29","ids":{"openalex":"https://openalex.org/W7166706237","doi":"https://doi.org/10.48550/arxiv.2606.29808"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.29808","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.29808","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.29808","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5121999607","display_name":"Y He","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Yuchen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133446682","display_name":"Peizhi Ying","orcid":"https://orcid.org/0009-0009-2494-8131"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ying, Peizhi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5109686566","display_name":"Liqi Cheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cheng, Liqi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133451444","display_name":"Kuilin Peng","orcid":"https://orcid.org/0009-0003-0945-3989"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Peng, Kuilin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100600477","display_name":"Yi Tian","orcid":"https://orcid.org/0000-0002-0831-6275"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tian, Yuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139671516","display_name":"Dazhen Deng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Deng, Dazhen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139677350","display_name":"Yingcai Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Yingcai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10601","display_name":"Handwritten Text Recognition Techniques","score":0.7990000247955322,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10601","display_name":"Handwritten Text Recognition Techniques","score":0.7990000247955322,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T11719","display_name":"Data Quality and Management","score":0.0364999994635582,"subfield":{"id":"https://openalex.org/subfields/1803","display_name":"Management Science and Operations Research"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T10799","display_name":"Data Visualization and Analytics","score":0.025100000202655792,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/chart","display_name":"Chart","score":0.763700008392334},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.6944000124931335},{"id":"https://openalex.org/keywords/workflow","display_name":"Workflow","score":0.6110000014305115},{"id":"https://openalex.org/keywords/table","display_name":"Table (database)","score":0.5611000061035156},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.5084999799728394},{"id":"https://openalex.org/keywords/data-modeling","display_name":"Data modeling","score":0.3555000126361847},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.34700000286102295}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7763000130653381},{"id":"https://openalex.org/C190812933","wikidata":"https://www.wikidata.org/wiki/Q28923","display_name":"Chart","level":2,"score":0.763700008392334},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.6944000124931335},{"id":"https://openalex.org/C177212765","wikidata":"https://www.wikidata.org/wiki/Q627335","display_name":"Workflow","level":2,"score":0.6110000014305115},{"id":"https://openalex.org/C45235069","wikidata":"https://www.wikidata.org/wiki/Q278425","display_name":"Table (database)","level":2,"score":0.5611000061035156},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.513700008392334},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5084999799728394},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.48500001430511475},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.39750000834465027},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.3555000126361847},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.34700000286102295},{"id":"https://openalex.org/C2777466982","wikidata":"https://www.wikidata.org/wiki/Q5227287","display_name":"Data extraction","level":3,"score":0.33559998869895935},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.3212999999523163},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.28850001096725464},{"id":"https://openalex.org/C113843644","wikidata":"https://www.wikidata.org/wiki/Q901882","display_name":"Interface (matter)","level":4,"score":0.288100004196167},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.28189998865127563},{"id":"https://openalex.org/C48677424","wikidata":"https://www.wikidata.org/wiki/Q6888088","display_name":"Mode (computer interface)","level":2,"score":0.2596000134944916}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.29808","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.29808","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.29808","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.29808","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","score":0.4501607120037079,"display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Chart":[0],"data":[1,5,52,69,97,148],"extraction,":[2],"which":[3],"reverse-engineers":[4],"tables":[6],"from":[7,99],"chart":[8,44,96,147],"images,":[9],"is":[10],"essential":[11],"for":[12,43,145],"reproducibility,":[13],"analysis,":[14],"retrieval,":[15],"and":[16,25,103],"redesign.":[17],"Existing":[18],"interactive":[19],"tools":[20],"are":[21],"reliable":[22,146],"but":[23],"tedious,":[24],"mixed-initiative":[26,143],"systems,":[27],"while":[28,78],"more":[29],"efficient,":[30],"lack":[31],"generalizability.":[32],"Recent":[33],"multimodal":[34],"large":[35],"language":[36],"models":[37],"(MLLMs)":[38],"offer":[39],"a":[40,62,100,109,130],"unified":[41],"interface":[42],"interpretation,":[45],"yet":[46],"their":[47],"ability":[48],"to":[49,71,114],"extract":[50],"accurate":[51],"tables,":[53],"especially":[54],"without":[55,68],"visible":[56],"labels,":[57],"remains":[58],"unclear.":[59],"We":[60],"build":[61],"benchmark":[63],"featuring":[64],"diverse":[65],"real-world":[66],"charts":[67],"labels":[70],"evaluate":[72],"this":[73],"capability.":[74],"Results":[75],"show":[76],"that,":[77],"current":[79],"MLLMs":[80],"reliably":[81],"reconstruct":[82],"table":[83],"structures,":[84],"they":[85],"struggle":[86],"with":[87,129],"precise":[88],"value":[89],"recovery.":[90],"To":[91],"address":[92],"this,":[93],"we":[94],"revisit":[95],"extraction":[98,106],"human-centered":[101],"perspective":[102],"argue":[104],"that":[105,138],"should":[107],"follow":[108],"progressive":[110],"learning":[111],"process":[112],"similar":[113],"how":[115],"people":[116],"read":[117],"charts.":[118],"Our":[119],"training":[120],"framework":[121],"substantially":[122],"improves":[123],"numerical":[124],"accuracy,":[125],"achieving":[126],"state-of-the-art":[127],"performance":[128],"7B-parameter":[131],"model.":[132],"A":[133],"user":[134],"study":[135],"further":[136],"shows":[137],"our":[139],"model":[140],"effectively":[141],"supports":[142],"workflows":[144],"extraction.":[149]},"counts_by_year":[],"updated_date":"2026-07-01T06:29:00.853634","created_date":"2026-07-01T00:00:00"}