{"id":"https://openalex.org/W7131106748","doi":"https://doi.org/10.1145/3795527","title":"Direct Retrieval-augmented Optimization: Synergizing Knowledge Selection and Language Models","display_name":"Direct Retrieval-augmented Optimization: Synergizing Knowledge Selection and Language Models","publication_year":2026,"publication_date":"2026-02-23","ids":{"openalex":"https://openalex.org/W7131106748","doi":"https://doi.org/10.1145/3795527"},"language":"en","primary_location":{"id":"doi:10.1145/3795527","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3795527","pdf_url":null,"source":{"id":"https://openalex.org/S4394735545","display_name":"ACM Transactions on Information Systems","issn_l":"1046-8188","issn":["1046-8188","1558-2868"],"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Information Systems","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Zhengliang Shi","orcid":"https://orcid.org/0000-0002-9658-4906"},"institutions":[{"id":"https://openalex.org/I121797337","display_name":"Leiden University","ror":"https://ror.org/027bh9e22","country_code":"NL","type":"education","lineage":["https://openalex.org/I121797337"]}],"countries":["NL"],"is_corresponding":false,"raw_author_name":"Zhengliang Shi","raw_affiliation_strings":["Leiden University, Leiden, The Netherlands"],"raw_orcid":"https://orcid.org/0000-0002-9658-4906","affiliations":[{"raw_affiliation_string":"Leiden University, Leiden, The Netherlands","institution_ids":["https://openalex.org/I121797337"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5050161084","display_name":"Lingyong Yan","orcid":"https://orcid.org/0000-0002-6547-1984"},"institutions":[{"id":"https://openalex.org/I98301712","display_name":"Baidu (China)","ror":"https://ror.org/03vs3wt56","country_code":"CN","type":"company","lineage":["https://openalex.org/I98301712"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Lingyong Yan","raw_affiliation_strings":["Baidu Inc., Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-6547-1984","affiliations":[{"raw_affiliation_string":"Baidu Inc., Beijing, China","institution_ids":["https://openalex.org/I98301712"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100662254","display_name":"Weiwei Sun","orcid":"https://orcid.org/0000-0002-4817-9500"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Weiwei Sun","raw_affiliation_strings":["LTI, Carnegie Mellon University, Pittsburgh, Pennsylvania, USA"],"raw_orcid":"https://orcid.org/0000-0002-4817-9500","affiliations":[{"raw_affiliation_string":"LTI, Carnegie Mellon University, Pittsburgh, Pennsylvania, USA","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126612943","display_name":"Yue Feng","orcid":null},"institutions":[{"id":"https://openalex.org/I79619799","display_name":"University of Birmingham","ror":"https://ror.org/03angcq70","country_code":"GB","type":"education","lineage":["https://openalex.org/I79619799"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Yue Feng","raw_affiliation_strings":["School of Computer Science, University of Birmingham, Birmingham, UK"],"raw_orcid":"https://orcid.org/0009-0000-6665-6406","affiliations":[{"raw_affiliation_string":"School of Computer Science, University of Birmingham, Birmingham, UK","institution_ids":["https://openalex.org/I79619799"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Pengjie Ren","orcid":"https://orcid.org/0000-0003-2964-6422"},"institutions":[{"id":"https://openalex.org/I154099455","display_name":"Shandong University","ror":"https://ror.org/0207yh398","country_code":"CN","type":"education","lineage":["https://openalex.org/I154099455"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Pengjie Ren","raw_affiliation_strings":["Shandong University, Qingdao, China"],"raw_orcid":"https://orcid.org/0000-0003-2964-6422","affiliations":[{"raw_affiliation_string":"Shandong University, Qingdao, China","institution_ids":["https://openalex.org/I154099455"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126648245","display_name":"Xinyu Ma","orcid":null},"institutions":[{"id":"https://openalex.org/I98301712","display_name":"Baidu (China)","ror":"https://ror.org/03vs3wt56","country_code":"CN","type":"company","lineage":["https://openalex.org/I98301712"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xinyu Ma","raw_affiliation_strings":["Baidu Inc., Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-5511-9370","affiliations":[{"raw_affiliation_string":"Baidu Inc., Beijing, China","institution_ids":["https://openalex.org/I98301712"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126609763","display_name":"Shuaiqiang Wang","orcid":null},"institutions":[{"id":"https://openalex.org/I98301712","display_name":"Baidu (China)","ror":"https://ror.org/03vs3wt56","country_code":"CN","type":"company","lineage":["https://openalex.org/I98301712"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shuaiqiang Wang","raw_affiliation_strings":["Baidu Inc., Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-9212-1947","affiliations":[{"raw_affiliation_string":"Baidu Inc., Beijing, China","institution_ids":["https://openalex.org/I98301712"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126616010","display_name":"Dawei Yin","orcid":null},"institutions":[{"id":"https://openalex.org/I98301712","display_name":"Baidu (China)","ror":"https://ror.org/03vs3wt56","country_code":"CN","type":"company","lineage":["https://openalex.org/I98301712"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Dawei Yin","raw_affiliation_strings":["Baidu Inc., Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-0684-6205","affiliations":[{"raw_affiliation_string":"Baidu Inc., Beijing, China","institution_ids":["https://openalex.org/I98301712"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126602511","display_name":"Maarten de Rijke","orcid":null},"institutions":[{"id":"https://openalex.org/I887064364","display_name":"University of Amsterdam","ror":"https://ror.org/04dkp9463","country_code":"NL","type":"education","lineage":["https://openalex.org/I887064364"]}],"countries":["NL"],"is_corresponding":false,"raw_author_name":"Maarten de Rijke","raw_affiliation_strings":["University of Amsterdam, Amsterdam, The Netherlands"],"raw_orcid":"https://orcid.org/0000-0002-1086-0202","affiliations":[{"raw_affiliation_string":"University of Amsterdam, Amsterdam, The Netherlands","institution_ids":["https://openalex.org/I887064364"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5126624352","display_name":"Zhaochun Ren","orcid":null},"institutions":[{"id":"https://openalex.org/I121797337","display_name":"Leiden University","ror":"https://ror.org/027bh9e22","country_code":"NL","type":"education","lineage":["https://openalex.org/I121797337"]}],"countries":["NL"],"is_corresponding":false,"raw_author_name":"Zhaochun Ren","raw_affiliation_strings":["Leiden University, Leiden, The Netherlands"],"raw_orcid":"https://orcid.org/0000-0002-9076-6565","affiliations":[{"raw_affiliation_string":"Leiden University, Leiden, The Netherlands","institution_ids":["https://openalex.org/I121797337"]}]}],"institutions":[],"countries_distinct_count":4,"institutions_distinct_count":6,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.18032497,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":95},"biblio":{"volume":"44","issue":"4","first_page":"1","last_page":"30"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.39100000262260437,"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/T10028","display_name":"Topic Modeling","score":0.39100000262260437,"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/T10286","display_name":"Information Retrieval and Search Behavior","score":0.23090000450611115,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.1242000013589859,"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/selection","display_name":"Selection (genetic algorithm)","score":0.6579999923706055},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.6395999789237976},{"id":"https://openalex.org/keywords/maximization","display_name":"Maximization","score":0.6276000142097473},{"id":"https://openalex.org/keywords/permutation","display_name":"Permutation (music)","score":0.4675000011920929},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.4652000069618225},{"id":"https://openalex.org/keywords/latent-variable","display_name":"Latent variable","score":0.4514000117778778},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.4374000132083893},{"id":"https://openalex.org/keywords/variance","display_name":"Variance (accounting)","score":0.43720000982284546},{"id":"https://openalex.org/keywords/expectation\u2013maximization-algorithm","display_name":"Expectation\u2013maximization algorithm","score":0.4172999858856201}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8580999970436096},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.6579999923706055},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.6395999789237976},{"id":"https://openalex.org/C2776330181","wikidata":"https://www.wikidata.org/wiki/Q18358244","display_name":"Maximization","level":2,"score":0.6276000142097473},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.6054999828338623},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5909000039100647},{"id":"https://openalex.org/C21308566","wikidata":"https://www.wikidata.org/wiki/Q7169365","display_name":"Permutation (music)","level":2,"score":0.4675000011920929},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.4652000069618225},{"id":"https://openalex.org/C51167844","wikidata":"https://www.wikidata.org/wiki/Q4422623","display_name":"Latent variable","level":2,"score":0.4514000117778778},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.4374000132083893},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.43720000982284546},{"id":"https://openalex.org/C182081679","wikidata":"https://www.wikidata.org/wiki/Q1275153","display_name":"Expectation\u2013maximization algorithm","level":3,"score":0.4172999858856201},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.4106000065803528},{"id":"https://openalex.org/C93959086","wikidata":"https://www.wikidata.org/wiki/Q6888345","display_name":"Model selection","level":2,"score":0.39879998564720154},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.38109999895095825},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.3700999915599823},{"id":"https://openalex.org/C61224824","wikidata":"https://www.wikidata.org/wiki/Q2260434","display_name":"Mixture model","level":2,"score":0.29100000858306885},{"id":"https://openalex.org/C9357733","wikidata":"https://www.wikidata.org/wiki/Q6878417","display_name":"Missing data","level":2,"score":0.290800005197525},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.28029999136924744},{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.2793000042438507},{"id":"https://openalex.org/C190839683","wikidata":"https://www.wikidata.org/wiki/Q2448197","display_name":"Train","level":2,"score":0.26489999890327454},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.2632000148296356},{"id":"https://openalex.org/C148483581","wikidata":"https://www.wikidata.org/wiki/Q446488","display_name":"Feature selection","level":2,"score":0.2567000091075897},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.25519999861717224},{"id":"https://openalex.org/C49781872","wikidata":"https://www.wikidata.org/wiki/Q1045555","display_name":"Maximum likelihood","level":2,"score":0.2547999918460846},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.25110000371932983}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3795527","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3795527","pdf_url":null,"source":{"id":"https://openalex.org/S4394735545","display_name":"ACM Transactions on Information Systems","issn_l":"1046-8188","issn":["1046-8188","1558-2868"],"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Information Systems","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G2522204841","display_name":null,"funder_award_id":"024.004.022, NWA.1389.20.183, and KICH3.LTP.20.006","funder_id":"https://openalex.org/F4320321800","funder_display_name":"Nederlandse Organisatie voor Wetenschappelijk Onderzoek"}],"funders":[{"id":"https://openalex.org/F4320321800","display_name":"Nederlandse Organisatie voor Wetenschappelijk Onderzoek","ror":"https://ror.org/04jsz6e67"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":11,"referenced_works":["https://openalex.org/W1985690171","https://openalex.org/W2050798338","https://openalex.org/W3099700870","https://openalex.org/W3100107515","https://openalex.org/W3190126809","https://openalex.org/W3217305727","https://openalex.org/W4384652670","https://openalex.org/W4386566442","https://openalex.org/W4400530533","https://openalex.org/W4401857375","https://openalex.org/W4405144021"],"related_works":[],"abstract_inverted_index":{"Retrieval-augmented":[0,100],"Generation":[1],"(RAG)":[2],"integrates":[3],"Large":[4],"Language":[5],"Models":[6],"(LLMs)":[7],"with":[8,221],"retrievers":[9],"to":[10,35,37,44,202],"access":[11],"external":[12],"knowledge,":[13],"improving":[14,139],"the":[15,25,33,42,88,147,164,174,179,188,218,232],"factuality":[16],"of":[17,72,110,237],"LLM":[18,123,192],"generation":[19],"in":[20,205,224],"knowledge-grounded":[21],"tasks.":[22],"To":[23],"optimize":[24],"RAG":[26,90,140],"performance,":[27],"most":[28],"previous":[29],"work":[30,56],"independently":[31],"fine-tunes":[32],"retriever":[34],"adapt":[36],"frozen":[38],"LLMs":[39,43],"or":[40],"trains":[41],"use":[45],"documents":[46],"retrieved":[47],"by":[48,60,167],"off-the-shelf":[49],"retrievers,":[50],"lacking":[51],"end-to-end":[52,108],"training":[53,62,109],"supervision.":[54],"Recent":[55],"addresses":[57],"this":[58],"limitation":[59],"jointly":[61,186],"these":[63],"two":[64,111,128],"components":[65,141],"but":[66],"relies":[67],"on":[68,211,242],"overly":[69],"simplifying":[70],"assumptions":[71],"document":[73,131,152],"independence,":[74],"which":[75],"has":[76],"been":[77],"criticized":[78],"for":[79],"being":[80],"far":[81],"from":[82,163],"real-world":[83],"scenarios.":[84],"Thus,":[85],"effectively":[86],"optimizing":[87],"overall":[89],"performance":[91],"remains":[92],"a":[93,98,115,143,155],"critical":[94],"challenge.":[95],"We":[96,228],"propose":[97],"Direct":[99],"Optimization":[101],"(":[102],"DRO":[103,125,199,216,243],")":[104],"framework":[105],"that":[106,198,215],"enables":[107],"key":[112],"components:":[113],"(i)":[114,130],"generative":[116],"knowledge":[117],"selection":[118,165,189],"model":[119,166,190],"and":[120,134,158,185,191,226,235],"(ii)":[121,135],"an":[122,169],"generator.":[124,193],"alternates":[126],"between":[127],"phases:":[129],"permutation":[132,153],"estimation":[133,148],"re-weighted":[136],"maximization,":[137],"progressively":[138],"through":[142],"variational":[144],"approach.":[145],"In":[146,173],"step,":[149,176],"we":[150,177],"treat":[151],"as":[154],"latent":[156],"variable":[157],"directly":[159],"estimate":[160],"its":[161],"distribution":[162],"applying":[168],"importance":[170,183],"sampling":[171],"strategy.":[172],"maximization":[175],"calibrate":[178],"optimization":[180],"expectation":[181],"using":[182],"weights":[184],"train":[187],"Our":[194],"theoretical":[195],"analysis":[196],"reveals":[197],"is":[200,240],"analogous":[201],"policy-gradient":[203],"methods":[204],"reinforcement":[206],"learning.":[207],"Extensive":[208],"experiments":[209],"conducted":[210],"five":[212],"datasets":[213],"illustrate":[214],"outperforms":[217],"best":[219],"baseline":[220],"5\u201315%":[222],"improvements":[223],"EM":[225],"F1.":[227],"also":[229],"qualitatively":[230],"analyze":[231],"stability,":[233],"convergence,":[234],"variance":[236],"DRO.":[238],"(Code":[239],"available":[241],"GitHub":[244],").":[245]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-05-21T06:26:12.895304","created_date":"2026-02-24T00:00:00"}