{"id":"https://openalex.org/W7163988287","doi":"https://doi.org/10.48550/arxiv.2606.07778","title":"Unlocking Latent Value: Taxonomy-Guided Recovery of High-Performing Data from Low-Tier Web Corpora","display_name":"Unlocking Latent Value: Taxonomy-Guided Recovery of High-Performing Data from Low-Tier Web Corpora","publication_year":2026,"publication_date":"2026-06-05","ids":{"openalex":"https://openalex.org/W7163988287","doi":"https://doi.org/10.48550/arxiv.2606.07778"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.07778","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.07778","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.2606.07778","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5138220645","display_name":"Neeraj Varshney","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Varshney, Neeraj","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5116225215","display_name":"Sanket Lokegaonkar","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lokegaonkar, Sanket","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089887413","display_name":"Nasser Zalmout","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zalmout, Nasser","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100610166","display_name":"Qingyu Yin","orcid":"https://orcid.org/0009-0006-1129-4704"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yin, Qingyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5007202198","display_name":"Priyanka Nigam","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Nigam, Priyanka","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5138214877","display_name":"Bing Yin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yin, Bing","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/T12016","display_name":"Web Data Mining and Analysis","score":0.7006999850273132,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T12016","display_name":"Web Data Mining and Analysis","score":0.7006999850273132,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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":0.16609999537467957,"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/T13976","display_name":"Web visibility and informetrics","score":0.032999999821186066,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/inference","display_name":"Inference","score":0.692799985408783},{"id":"https://openalex.org/keywords/pairwise-comparison","display_name":"Pairwise comparison","score":0.5885999798774719},{"id":"https://openalex.org/keywords/filter","display_name":"Filter (signal processing)","score":0.5516999959945679},{"id":"https://openalex.org/keywords/coding","display_name":"Coding (social sciences)","score":0.45980000495910645},{"id":"https://openalex.org/keywords/dimension","display_name":"Dimension (graph theory)","score":0.4388999938964844},{"id":"https://openalex.org/keywords/missing-data","display_name":"Missing data","score":0.35030001401901245}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7469000220298767},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.692799985408783},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.5885999798774719},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.5554999709129333},{"id":"https://openalex.org/C106131492","wikidata":"https://www.wikidata.org/wiki/Q3072260","display_name":"Filter (signal processing)","level":2,"score":0.5516999959945679},{"id":"https://openalex.org/C179518139","wikidata":"https://www.wikidata.org/wiki/Q5140297","display_name":"Coding (social sciences)","level":2,"score":0.45980000495910645},{"id":"https://openalex.org/C33676613","wikidata":"https://www.wikidata.org/wiki/Q13415176","display_name":"Dimension (graph theory)","level":2,"score":0.4388999938964844},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.40049999952316284},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.36079999804496765},{"id":"https://openalex.org/C9357733","wikidata":"https://www.wikidata.org/wiki/Q6878417","display_name":"Missing data","level":2,"score":0.35030001401901245},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3073999881744385},{"id":"https://openalex.org/C24756922","wikidata":"https://www.wikidata.org/wiki/Q1757694","display_name":"Data quality","level":3,"score":0.2987000048160553},{"id":"https://openalex.org/C118643609","wikidata":"https://www.wikidata.org/wiki/Q189210","display_name":"Web application","level":2,"score":0.2980000078678131},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.2768999934196472},{"id":"https://openalex.org/C2776291640","wikidata":"https://www.wikidata.org/wiki/Q2912517","display_name":"Value (mathematics)","level":2,"score":0.27160000801086426},{"id":"https://openalex.org/C136085584","wikidata":"https://www.wikidata.org/wiki/Q910289","display_name":"Overlay","level":2,"score":0.25380000472068787}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.07778","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.07778","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.2606.07778","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.07778","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":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Dominant":[0],"web":[1,159,249],"data":[2,199,210,234],"curation":[3],"pipelines":[4],"for":[5],"pretraining":[6],"collapse":[7],"document":[8],"quality":[9],"into":[10,79],"a":[11,26,80,92,114,147,257],"single":[12],"composite":[13,40],"score,":[14],"systematically":[15],"missing":[16],"high-value":[17],"content":[18],"along":[19,35],"dimensions":[20,38],"the":[21,48,106,121,139,154,170,215],"scorer":[22],"underweights.":[23],"We":[24,70],"present":[25],"taxonomy-driven":[27],"framework":[28],"that":[29,39,241,252],"recovers":[30],"this":[31],"value":[32,244],"by":[33,184,200,220],"filtering":[34,255],"semantically":[36],"meaningful":[37],"scores":[41],"fail":[42],"to":[43,104,157,261],"capture.":[44],"First,":[45],"building":[46],"on":[47,96,186,189,193,202,206,222,226,235],"ESSENTIAL-WEB":[49],"taxonomy,":[50],"we":[51,89,112],"introduce":[52,113],"two":[53,212],"novel":[54],"dimensions:":[55],"timeliness":[56],"and":[57,77,131,134,167,191,204,224,251],"cultural":[58],"specificity,":[59],"both":[60],"of":[61,109,149],"which":[62],"show":[63],"low":[64],"pairwise":[65],"NMI":[66],"with":[67],"existing":[68],"ones.":[69],"annotate":[71],"14M":[72],"documents":[73],"using":[74],"Qwen2.5":[75],"32B":[76],"distill":[78],"lightweight":[81],"0.5B":[82],"model.":[83],"To":[84],"enable":[85],"rapid":[86],"corpus-wide":[87],"annotation,":[88],"additionally":[90],"train":[91],"73M":[93],"multi-task":[94],"MLP":[95],"E5":[97],"embeddings,":[98],"achieving":[99],"50x":[100],"inference":[101],"throughput.":[102],"Second,":[103],"navigate":[105],"combinatorial":[107],"explosion":[108],"filter":[110,178],"configurations,":[111],"compute-efficient":[115,259],"two-pass":[116],"framework:":[117],"Pass":[118,128],"1":[119],"identifies":[120],"strongest":[122],"dimension":[123],"signals":[124],"at":[125,146],"small":[126],"scale;":[127],"2":[129],"constructs":[130],"evaluates":[132],"conjunctive":[133],"disjunctive":[135],"compound":[136],"filters":[137,156],"from":[138,211],"top":[140],"performers":[141],"-":[142],"identifying":[143],"high-performing":[144],"configurations":[145],"fraction":[148],"full":[150],"scaling-law":[151],"cost.":[152],"Applying":[153],"selected":[155],"deprioritized":[158,248],"data,":[160,175,250],"taxonomy-filtered":[161],"subsets":[162],"outperform":[163],"their":[164],"unfiltered":[165,182,197,230],"baselines":[166],"even":[168],"surpass":[169],"highest-quality":[171],"tier.":[172],"On":[173],"mid-tier":[174],"our":[176],"best":[177],"improves":[179,219],"over":[180,228],"its":[181,229],"baseline":[183],"12.1%":[185],"reasoning,":[187],"9.5%":[188],"coding,":[190],"2.0%":[192],"knowledge":[194],"benchmarks,":[195],"exceeding":[196],"top-tier":[198,233],"6.7%":[201],"reasoning":[203,223],"13.7%":[205],"coding.":[207],"Furthermore,":[208],"filtered":[209],"tiers":[213],"below":[214],"typical":[216],"production":[217],"threshold":[218],"22.3%":[221],"19.5%":[225],"coding":[227,236],"baseline,":[231],"surpassing":[232],"benchmarks.":[237],"These":[238],"results":[239],"establish":[240],"vast":[242],"latent":[243],"remains":[245],"locked":[246],"in":[247],"multi-dimensional":[253],"taxonomy":[254],"is":[256],"principled,":[258],"key":[260],"unlocking":[262],"it.":[263]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-10T00:00:00"}