{"id":"https://openalex.org/W7161280307","doi":"https://doi.org/10.48550/arxiv.2605.14434","title":"Efficient Generative Retrieval for E-commerce Search with Semantic Cluster IDs and Expert-Guided RL","display_name":"Efficient Generative Retrieval for E-commerce Search with Semantic Cluster IDs and Expert-Guided RL","publication_year":2026,"publication_date":"2026-05-14","ids":{"openalex":"https://openalex.org/W7161280307","doi":"https://doi.org/10.48550/arxiv.2605.14434"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.14434","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.14434","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","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.2605.14434","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5136242775","display_name":"Jianbo Zhu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhu, Jianbo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136197672","display_name":"Xing Fang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fang, Xing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136235554","display_name":"Jing Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Jing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5061017731","display_name":"Mingmin Jin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jin, Mingmin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136189129","display_name":"Bokang Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Bokang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136215736","display_name":"Guangxin Song","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song, Guangxin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136206822","display_name":"Zhenyu Xie","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xie, Zhenyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5101874862","display_name":"Junjie Bai","orcid":"https://orcid.org/0009-0003-6805-4785"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bai, Junjie","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/T10286","display_name":"Information Retrieval and Search Behavior","score":0.5062999725341797,"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/T10286","display_name":"Information Retrieval and Search Behavior","score":0.5062999725341797,"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/T10203","display_name":"Recommender Systems and Techniques","score":0.1565999984741211,"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/T13274","display_name":"Expert finding and Q&A systems","score":0.04479999840259552,"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/generative-grammar","display_name":"Generative grammar","score":0.5713000297546387},{"id":"https://openalex.org/keywords/ranking","display_name":"Ranking (information retrieval)","score":0.5601999759674072},{"id":"https://openalex.org/keywords/learning-to-rank","display_name":"Learning to rank","score":0.40139999985694885},{"id":"https://openalex.org/keywords/semantic-search","display_name":"Semantic search","score":0.3982999920845032},{"id":"https://openalex.org/keywords/precision-and-recall","display_name":"Precision and recall","score":0.3952000141143799},{"id":"https://openalex.org/keywords/recall","display_name":"Recall","score":0.3928999900817871},{"id":"https://openalex.org/keywords/document-retrieval","display_name":"Document retrieval","score":0.3871999979019165},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.38100001215934753}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7644000053405762},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.5713000297546387},{"id":"https://openalex.org/C189430467","wikidata":"https://www.wikidata.org/wiki/Q7293293","display_name":"Ranking (information retrieval)","level":2,"score":0.5601999759674072},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.5461000204086304},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5052000284194946},{"id":"https://openalex.org/C86037889","wikidata":"https://www.wikidata.org/wiki/Q4330127","display_name":"Learning to rank","level":3,"score":0.40139999985694885},{"id":"https://openalex.org/C166423231","wikidata":"https://www.wikidata.org/wiki/Q1891170","display_name":"Semantic search","level":3,"score":0.3982999920845032},{"id":"https://openalex.org/C81669768","wikidata":"https://www.wikidata.org/wiki/Q2359161","display_name":"Precision and recall","level":2,"score":0.3952000141143799},{"id":"https://openalex.org/C100660578","wikidata":"https://www.wikidata.org/wiki/Q18733","display_name":"Recall","level":2,"score":0.3928999900817871},{"id":"https://openalex.org/C161156560","wikidata":"https://www.wikidata.org/wiki/Q1638872","display_name":"Document retrieval","level":2,"score":0.3871999979019165},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.38100001215934753},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.3790000081062317},{"id":"https://openalex.org/C19889080","wikidata":"https://www.wikidata.org/wiki/Q2835852","display_name":"Beam search","level":3,"score":0.37709999084472656},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.3659999966621399},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.36039999127388},{"id":"https://openalex.org/C2776207758","wikidata":"https://www.wikidata.org/wiki/Q5303302","display_name":"Downstream (manufacturing)","level":2,"score":0.3601999878883362},{"id":"https://openalex.org/C66746571","wikidata":"https://www.wikidata.org/wiki/Q1134833","display_name":"ENCODE","level":3,"score":0.3091000020503998},{"id":"https://openalex.org/C127162648","wikidata":"https://www.wikidata.org/wiki/Q16858953","display_name":"Channel (broadcasting)","level":2,"score":0.2892000079154968},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.27709999680519104},{"id":"https://openalex.org/C90673727","wikidata":"https://www.wikidata.org/wiki/Q901718","display_name":"Product (mathematics)","level":2,"score":0.266400009393692},{"id":"https://openalex.org/C197914299","wikidata":"https://www.wikidata.org/wiki/Q18650","display_name":"Semantic memory","level":3,"score":0.26409998536109924},{"id":"https://openalex.org/C1667742","wikidata":"https://www.wikidata.org/wiki/Q10927554","display_name":"Image retrieval","level":3,"score":0.25690001249313354},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.2556999921798706},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2554999887943268}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.14434","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.14434","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.14434","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.14434","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Generative":[0],"retrieval":[1,11,43,54,63,210],"offers":[2],"a":[3,14,53,65,113,204],"promising":[4],"alternative":[5],"by":[6,127],"unifying":[7],"the":[8,29,39],"fragmented":[9],"multi-stage":[10],"process":[12],"into":[13,94],"single":[15],"end-to-end":[16,71],"model.":[17],"However,":[18],"its":[19],"practical":[20],"adoption":[21],"in":[22,151,175,188,211],"industrial":[23],"e-commerce":[24,213],"search":[25,102,138,163],"remains":[26],"challenging,":[27],"given":[28],"massive":[30],"and":[31,38,82,147,153,178,199],"dynamic":[32],"product":[33],"catalogs,":[34],"strict":[35],"latency":[36],"requirements,":[37],"need":[40],"to":[41,91,131,145],"align":[42],"with":[44,87,121],"downstream":[45,122],"ranking":[46,123],"goals.":[47],"In":[48],"this":[49],"work,":[50],"we":[51,105],"propose":[52],"framework":[55],"tailored":[56],"for":[57,191,207],"real-world":[58,212],"recall":[59,120,183],"scenarios,":[60],"positioning":[61],"generative":[62,119,182,209],"as":[64],"recall-stage":[66],"supplement":[67],"rather":[68],"than":[69],"an":[70],"replacement.":[72],"Our":[73],"method,":[74],"CQ-SID":[75,142],"(Category-and-Query":[76],"constrained":[77],"Semantic":[78],"ID),":[79],"employs":[80],"category-aware":[81],"query-item":[83],"contrastive":[84],"learning":[85,115],"along":[86],"Residual":[88],"Quantized":[89],"VAEs":[90],"encode":[92],"items":[93],"hierarchical":[95],"semantic":[96,152],"cluster":[97],"identifiers,":[98],"significantly":[99],"reducing":[100],"beam":[101,162],"complexity.":[103],"Additionally,":[104],"develop":[106],"EG-GRPO":[107,165],"(Expert-Guided":[108],"Group":[109],"Relative":[110],"Policy":[111],"Optimization),":[112],"reinforcement":[114],"approach":[116],"that":[117,141],"aligns":[118],"under":[124],"sparse":[125],"rewards":[126],"injecting":[128],"ground-truth":[129],"samples":[130],"stabilize":[132],"training.":[133],"Offline":[134],"experiments":[135],"on":[136],"TmallAPP":[137],"logs":[139],"show":[140],"achieves":[143],"up":[144],"26.76%":[146],"11.11%":[148],"relative":[149],"gains":[150,174],"personalized":[154],"click":[155],"hitrate":[156],"over":[157,192],"RQ-VAE":[158],"baselines,":[159],"while":[160],"halving":[161],"size.":[164],"further":[166],"improves":[167],"multi-objective":[168],"performance.":[169],"Online":[170],"A/B":[171],"tests":[172],"confirm":[173],"GMV":[176],"(+1.15%)":[177],"UCTCVR":[179],"(+0.40%).":[180],"The":[181],"channel":[184],"now":[185],"contributes":[186],"substantially":[187],"production,":[189],"accounting":[190],"50.25%":[193],"of":[194,197,201],"exposures,":[195],"58.96%":[196],"clicks,":[198],"72.63%":[200],"purchases,":[202],"demonstrating":[203],"viable":[205],"path":[206],"deploying":[208],"systems.":[214]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-16T00:00:00"}