{"id":"https://openalex.org/W7160112008","doi":"https://doi.org/10.1016/j.sysarc.2026.103783","title":"Preference-driven prompt refinement for software development: A cross-model analysis of GPT-5","display_name":"Preference-driven prompt refinement for software development: A cross-model analysis of GPT-5","publication_year":2026,"publication_date":"2026-04-30","ids":{"openalex":"https://openalex.org/W7160112008","doi":"https://doi.org/10.1016/j.sysarc.2026.103783"},"language":"en","primary_location":{"id":"doi:10.1016/j.sysarc.2026.103783","is_oa":true,"landing_page_url":"https://doi.org/10.1016/j.sysarc.2026.103783","pdf_url":null,"source":{"id":"https://openalex.org/S127660348","display_name":"Journal of Systems Architecture","issn_l":"1383-7621","issn":["1383-7621","1873-6165"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320990","host_organization_name":"Elsevier BV","host_organization_lineage":["https://openalex.org/P4310320990"],"host_organization_lineage_names":["Elsevier BV"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Journal of Systems Architecture","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://doi.org/10.1016/j.sysarc.2026.103783","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5042751684","display_name":"Ashraf Elnashar","orcid":null},"institutions":[{"id":"https://openalex.org/I200719446","display_name":"Vanderbilt University","ror":"https://ror.org/02vm5rt34","country_code":"US","type":"education","lineage":["https://openalex.org/I200719446"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Ashraf Elnashar","raw_affiliation_strings":["Department of Computer Science, Vanderbilt University, Nashville, TN, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science, Vanderbilt University, Nashville, TN, USA","institution_ids":["https://openalex.org/I200719446"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023026501","display_name":"Jules White","orcid":"https://orcid.org/0000-0002-6331-2365"},"institutions":[{"id":"https://openalex.org/I200719446","display_name":"Vanderbilt University","ror":"https://ror.org/02vm5rt34","country_code":"US","type":"education","lineage":["https://openalex.org/I200719446"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jules White","raw_affiliation_strings":["Department of Computer Science, Vanderbilt University, Nashville, TN, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science, Vanderbilt University, Nashville, TN, USA","institution_ids":["https://openalex.org/I200719446"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5135122613","display_name":"Douglas C. Schmidt","orcid":null},"institutions":[{"id":"https://openalex.org/I16285277","display_name":"William & Mary","ror":"https://ror.org/03hsf0573","country_code":"US","type":"education","lineage":["https://openalex.org/I16285277"]},{"id":"https://openalex.org/I267592682","display_name":"Williams (United States)","ror":"https://ror.org/007zhvp17","country_code":"US","type":"company","lineage":["https://openalex.org/I267592682"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Douglas C. Schmidt","raw_affiliation_strings":["Department of Computer Science, William & Mary, Williamsburg, VA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science, William & Mary, Williamsburg, VA, USA","institution_ids":["https://openalex.org/I16285277","https://openalex.org/I267592682"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5042751684"],"corresponding_institution_ids":["https://openalex.org/I200719446"],"apc_list":{"value":2550,"currency":"USD","value_usd":2550},"apc_paid":{"value":2550,"currency":"USD","value_usd":2550},"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.55095164,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"176","issue":null,"first_page":"103783","last_page":"103783"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10430","display_name":"Software Engineering Techniques and Practices","score":0.18140000104904175,"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/T10430","display_name":"Software Engineering Techniques and Practices","score":0.18140000104904175,"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/T10260","display_name":"Software Engineering Research","score":0.10320000350475311,"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/T10639","display_name":"Advanced Software Engineering Methodologies","score":0.08299999684095383,"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/software","display_name":"Software","score":0.54830002784729},{"id":"https://openalex.org/keywords/software-development","display_name":"Software development","score":0.2745000123977661},{"id":"https://openalex.org/keywords/software-system","display_name":"Software system","score":0.26980000734329224},{"id":"https://openalex.org/keywords/abstract-interpretation","display_name":"Abstract interpretation","score":0.26579999923706055}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.9121999740600586},{"id":"https://openalex.org/C2777904410","wikidata":"https://www.wikidata.org/wiki/Q7397","display_name":"Software","level":2,"score":0.54830002784729},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.5311999917030334},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.4065000116825104},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.30399999022483826},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.28769999742507935},{"id":"https://openalex.org/C529173508","wikidata":"https://www.wikidata.org/wiki/Q638608","display_name":"Software development","level":3,"score":0.2745000123977661},{"id":"https://openalex.org/C149091818","wikidata":"https://www.wikidata.org/wiki/Q2429814","display_name":"Software system","level":3,"score":0.26980000734329224},{"id":"https://openalex.org/C2780654840","wikidata":"https://www.wikidata.org/wiki/Q333341","display_name":"Abstract interpretation","level":2,"score":0.26579999923706055},{"id":"https://openalex.org/C2984968299","wikidata":"https://www.wikidata.org/wiki/Q1077784","display_name":"Software tool","level":3,"score":0.2402999997138977}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1016/j.sysarc.2026.103783","is_oa":true,"landing_page_url":"https://doi.org/10.1016/j.sysarc.2026.103783","pdf_url":null,"source":{"id":"https://openalex.org/S127660348","display_name":"Journal of Systems Architecture","issn_l":"1383-7621","issn":["1383-7621","1873-6165"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320990","host_organization_name":"Elsevier BV","host_organization_lineage":["https://openalex.org/P4310320990"],"host_organization_lineage_names":["Elsevier BV"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Journal of Systems Architecture","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1016/j.sysarc.2026.103783","is_oa":true,"landing_page_url":"https://doi.org/10.1016/j.sysarc.2026.103783","pdf_url":null,"source":{"id":"https://openalex.org/S127660348","display_name":"Journal of Systems Architecture","issn_l":"1383-7621","issn":["1383-7621","1873-6165"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320990","host_organization_name":"Elsevier BV","host_organization_lineage":["https://openalex.org/P4310320990"],"host_organization_lineage_names":["Elsevier BV"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Journal of Systems Architecture","raw_type":"journal-article"},"sustainable_development_goals":[{"score":0.4007928967475891,"display_name":"Industry, innovation and infrastructure","id":"https://metadata.un.org/sdg/9"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":1,"referenced_works":["https://openalex.org/W4396833223"],"related_works":[],"abstract_inverted_index":{"Rapid":[0],"advances":[1,324],"in":[2,59,280],"large":[3],"language":[4],"models":[5],"(LLMs)":[6],"have":[7],"expanded":[8],"their":[9,309],"use":[10],"across":[11,121],"software":[12,123,345,359],"engineering":[13,18,41,124,327,346],"workflows,":[14],"yet":[15],"most":[16],"prompt":[17,40,326],"remains":[19],"ad":[20,173],"hoc":[21,174],"\u2014yielding":[22],"inconsistent":[23],"quality,":[24],"poor":[25],"transferability,":[26],"and":[27,33,51,67,82,104,106,126,154,164,222,229,290,306,315,320,336,348],"inefficient":[28],"iteration.":[29,302],"This":[30,266],"paper":[31,323],"introduces":[32],"evaluates":[34,99],"Preference-Driven":[35],"Refinement":[36],"(PDR),":[37],"a":[38,90,95,117],"structured":[39,112,197],"method":[42],"that":[43,257],"iteratively":[44],"updates":[45],"prompts":[46],"by":[47],"incorporating":[48],"user-selected":[49],"preferred":[50],"non-preferred":[52],"elements":[53],"from":[54],"model":[55,98],"outputs.":[56],"Building":[57],"on":[58,207,352],"context":[60,80],"learning,":[61],"PDR":[62,87,166,189,305],"uses":[63],"synthetic":[64],"example":[65],"generation":[66],"explicit":[68,215],"preference":[69],"incorporation":[70],"to":[71,136,172,196],"overcome":[72],"common":[73],"challenges,":[74],"such":[75],"as":[76,156],"instruction":[77],"blindness,":[78],"inadequate":[79],"capture,":[81],"alignment":[83],"drift.":[84],"We":[85],"extend":[86],"with":[88,131,311,344],"PDR+Critic,":[89,307],"multi-LLM":[91,340],"refinement":[92,113,169,186],"loop":[93],"where":[94,234],"separate":[96],"Critic":[97],"generated":[100],"outputs,":[101],"identifies":[102],"strengths":[103],"weaknesses,":[105],"recommends":[107],"improvements.":[108],"To":[109,140],"assess":[110],"these":[111],"methods,":[114],"we":[115,145],"conduct":[116],"controlled,":[118],"simulation-based":[119],"study":[120],"five":[122],"tasks":[125],"eight":[127],"personas":[128],"using":[129,149],"GPT-5,":[130],"decoding":[132],"parameters":[133],"held":[134],"constant":[135],"isolate":[137],"methodological":[138,162],"effects.":[139],"address":[141],"evaluation":[142,216,242,355],"circularity":[143],"concerns,":[144],"employ":[146],"multi-model":[147,241],"cross-validation":[148],"independent":[150,254],"LLM":[151,259],"families":[152,260],"(GPT-4o":[153],"Gemini-2.5-Pro)":[155],"evaluators":[157,231],"Our":[158,329],"results":[159],"reveal":[160,331],"both":[161,230,318],"contributions":[163],"challenges.":[165],"significantly":[167],"reduces":[168],"iterations":[170,296],"compared":[171],"prompting":[175],"(p":[176],"=":[177,248,251],"0.020,":[178],"FDR-corrected),":[179],"demonstrating":[180],"improved":[181],"convergence":[182],"efficiency":[183,293],"through":[184],"fewer":[185,295],"cycles.":[187],"However,":[188],"exhibits":[190],"longer":[191,298],"total":[192],"execution":[193,299],"time":[194,300],"due":[195],"processing":[198],"overhead":[199],"(Ad":[200],"Hoc:":[201],"67.5s,":[202],"PDR:":[203],"179.6s,":[204],"PDR+Critic:":[205],"209.1s":[206],"average).":[208],"PDR+Critic":[209],"shows":[210],"mixed":[211],"effectiveness:":[212],"while":[213],"adding":[214],"loops,":[217],"it":[218],"often":[219],"increases":[220],"runtime":[221],"iteration":[223],"count":[224],"without":[225],"consistent":[226],"quality":[227,264,271],"improvements,":[228],"identified":[232],"cases":[233],"critic":[235],"feedback":[236],"degraded":[237],"output":[238],"quality.":[239],"Critically,":[240],"revealed":[243],"low":[244],"inter-rater":[245],"agreement":[246],"(r":[247],"0.044,":[249],"p":[250],"0.794)":[252],"between":[253,334],"evaluators,":[255],"indicating":[256],"different":[258,263],"apply":[261],"fundamentally":[262],"standards.":[265],"finding":[267],"weakens":[268],"claims":[269],"about":[270,277],"improvements":[272],"but":[273,297],"contributes":[274],"important":[275],"evidence":[276],"current":[278],"limitations":[279],"LLM-as-judge":[281],"methodologies.":[282],"Objective":[283],"metrics":[284],"(iteration":[285],"counts,":[286],"timing)":[287],"remain":[288],"valid":[289],"demonstrate":[291],"PDR\u2019s":[292],"trade-offs:":[294],"per":[301],"By":[303],"formalizing":[304],"evaluating":[308],"performance":[310],"rigorous":[312],"cross-model":[313],"validation,":[314],"transparently":[316],"reporting":[317],"successes":[319],"limitations,":[321],"this":[322],"systematic":[325],"research.":[328],"findings":[330],"key":[332],"trade-offs":[333],"structure":[335],"efficiency,":[337],"document":[338],"how":[339],"Generator\u2013Critic":[341],"loops":[342],"interact":[343],"tasks,":[347],"motivate":[349],"future":[350],"work":[351],"standardized,":[353],"reliable":[354],"frameworks":[356],"for":[357],"LLM-augmented":[358],"development.":[360]},"counts_by_year":[],"updated_date":"2026-07-23T08:03:31.855105","created_date":"2026-05-05T00:00:00"}