{"id":"https://openalex.org/W7162126093","doi":"https://doi.org/10.48550/arxiv.2605.21752","title":"PEARL: Unbiased Percentile Estimation via Contrastive Learning for Industrial-Scale Livestream Recommendation","display_name":"PEARL: Unbiased Percentile Estimation via Contrastive Learning for Industrial-Scale Livestream Recommendation","publication_year":2026,"publication_date":"2026-05-20","ids":{"openalex":"https://openalex.org/W7162126093","doi":"https://doi.org/10.48550/arxiv.2605.21752"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.21752","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.21752","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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.2605.21752","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5092675601","display_name":"Blake Gella","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gella, Blake","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136786102","display_name":"Wei Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Wei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136792669","display_name":"Yuhao Yin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yin, Yuhao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5044662028","display_name":"Zexi Huang","orcid":"https://orcid.org/0000-0002-1480-4494"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Zexi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136741947","display_name":"Zikai Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zikai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129680215","display_name":"Emily Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Emily","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136780397","display_name":"Junlin Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Junlin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124415930","display_name":"Wentao Guo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Wentao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136755848","display_name":"Qinglei Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Qinglei","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/T10203","display_name":"Recommender Systems and Techniques","score":0.9589999914169312,"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/T10203","display_name":"Recommender Systems and Techniques","score":0.9589999914169312,"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/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.007000000216066837,"subfield":{"id":"https://openalex.org/subfields/1706","display_name":"Computer Science Applications"},"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.003700000001117587,"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/percentile","display_name":"Percentile","score":0.6184999942779541},{"id":"https://openalex.org/keywords/recommender-system","display_name":"Recommender system","score":0.5633000135421753},{"id":"https://openalex.org/keywords/robustness","display_name":"Robustness (evolution)","score":0.5023000240325928},{"id":"https://openalex.org/keywords/smoothing","display_name":"Smoothing","score":0.4652999937534332},{"id":"https://openalex.org/keywords/pairwise-comparison","display_name":"Pairwise comparison","score":0.45159998536109924},{"id":"https://openalex.org/keywords/bootstrapping","display_name":"Bootstrapping (finance)","score":0.42980000376701355},{"id":"https://openalex.org/keywords/ranking","display_name":"Ranking (information retrieval)","score":0.4007999897003174},{"id":"https://openalex.org/keywords/resampling","display_name":"Resampling","score":0.3700000047683716},{"id":"https://openalex.org/keywords/collaborative-filtering","display_name":"Collaborative filtering","score":0.3481000065803528}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.690500020980835},{"id":"https://openalex.org/C122048520","wikidata":"https://www.wikidata.org/wiki/Q2913954","display_name":"Percentile","level":2,"score":0.6184999942779541},{"id":"https://openalex.org/C557471498","wikidata":"https://www.wikidata.org/wiki/Q554950","display_name":"Recommender system","level":2,"score":0.5633000135421753},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.5023000240325928},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.48739999532699585},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4666999876499176},{"id":"https://openalex.org/C3770464","wikidata":"https://www.wikidata.org/wiki/Q775963","display_name":"Smoothing","level":2,"score":0.4652999937534332},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.45159998536109924},{"id":"https://openalex.org/C207609745","wikidata":"https://www.wikidata.org/wiki/Q4944086","display_name":"Bootstrapping (finance)","level":2,"score":0.42980000376701355},{"id":"https://openalex.org/C189430467","wikidata":"https://www.wikidata.org/wiki/Q7293293","display_name":"Ranking (information retrieval)","level":2,"score":0.4007999897003174},{"id":"https://openalex.org/C150921843","wikidata":"https://www.wikidata.org/wiki/Q1170431","display_name":"Resampling","level":2,"score":0.3700000047683716},{"id":"https://openalex.org/C21569690","wikidata":"https://www.wikidata.org/wiki/Q94702","display_name":"Collaborative filtering","level":3,"score":0.3481000065803528},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3240000009536743},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.3095000088214874},{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.3046000003814697},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.29910001158714294},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.29319998621940613},{"id":"https://openalex.org/C102366305","wikidata":"https://www.wikidata.org/wiki/Q1097688","display_name":"Nonparametric statistics","level":2,"score":0.2824999988079071},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.2770000100135803},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.2667999863624573},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.2662999927997589},{"id":"https://openalex.org/C2776156558","wikidata":"https://www.wikidata.org/wiki/Q4353746","display_name":"MovieLens","level":4,"score":0.26339998841285706},{"id":"https://openalex.org/C2776502983","wikidata":"https://www.wikidata.org/wiki/Q690182","display_name":"Contrast (vision)","level":2,"score":0.26269999146461487},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.25949999690055847},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.25850000977516174},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.2533000111579895},{"id":"https://openalex.org/C149441793","wikidata":"https://www.wikidata.org/wiki/Q200726","display_name":"Probability distribution","level":2,"score":0.25200000405311584}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.21752","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.21752","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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.2605.21752","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.21752","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":{"Recommender":[0],"systems":[1],"trained":[2],"on":[3,100],"user":[4,184],"interaction":[5,91],"data":[6],"are":[7],"susceptible":[8],"to":[9,39,93,133,148],"behavioral":[10,164],"intensity":[11],"imbalance--a":[12],"systematic":[13],"distortion":[14],"arising":[15],"from":[16,43],"heterogeneous":[17],"engagement":[18,80],"patterns":[19],"across":[20,171],"users.":[21],"This":[22],"imbalance":[23],"skews":[24],"feedback":[25],"signals":[26,42,76],"such":[27,111],"that":[28,72,110,160],"observed":[29],"interactions":[30],"no":[31],"longer":[32],"faithfully":[33],"reflect":[34],"true":[35],"preferences,":[36],"causing":[37],"models":[38,73],"disproportionately":[40],"amplify":[41],"highly":[44],"active":[45],"users":[46],"while":[47],"underrepresenting":[48],"others,":[49],"which":[50],"ultimately":[51],"degrades":[52],"recommendation":[53,169],"quality":[54],"and":[55,136,144,153,166,204],"robustness":[56],"at":[57],"scale.":[58],"To":[59],"address":[60],"this":[61],"issue,":[62],"we":[63,124],"propose":[64],"a":[65,126,140,145,177,182],"nonparametric":[66],"contrastive":[67,90],"percentile":[68,95,131],"approximation":[69],"framework,":[70],"PEARL,":[71],"relative":[74,84],"preference":[75,119],"instead":[77],"of":[78,117,186],"absolute":[79],"magnitudes.":[81],"Building":[82],"upon":[83],"advantage":[85],"debiasing,":[86],"PEARL":[87,161],"leverages":[88],"real":[89],"samples":[92],"approximate":[94],"relationships":[96],"directly,":[97],"without":[98],"relying":[99],"auxiliary":[101],"distribution":[102],"estimation":[103],"models.":[104],"We":[105],"provide":[106],"theoretical":[107],"justification":[108],"demonstrating":[109],"pairwise":[112],"comparisons":[113],"yield":[114],"unbiased":[115],"estimates":[116],"percentile-based":[118],"signals.":[120],"For":[121],"broader":[122],"applicability,":[123],"introduce":[125],"prediction-based":[127],"bootstrapping":[128],"mechanism":[129],"for":[130],"smoothing":[132],"handle":[134],"sparse":[135],"discrete":[137],"feedback,":[138],"alongside":[139],"generalized":[141],"value-weighted":[142],"formulation":[143],"co-training":[146],"strategy":[147],"enhance":[149],"both":[150],"modeling":[151],"flexibility":[152],"representation":[154],"learning.":[155],"Extensive":[156],"offline":[157],"experiments":[158],"demonstrate":[159],"effectively":[162],"mitigates":[163],"bias":[165],"consistently":[167],"improves":[168],"performance":[170],"multiple":[172],"ranking":[173],"targets.":[174],"Deployed":[175],"in":[176],"production":[178],"livestream":[179],"platform":[180],"with":[181],"combined":[183],"base":[185],"billions,":[187],"online":[188],"A/B":[189],"testing":[190],"confirms":[191],"substantial":[192],"real-world":[193],"gains:":[194],"+2.10%":[195],"Watch":[196],"Duration,":[197],"+0.80%":[198],"Consumption":[199],"Amount,":[200],"+1.49%":[201],"Interaction":[202],"Rate,":[203],"-6.91%":[205],"Report":[206],"Rate.":[207]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-23T00:00:00"}