[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85858-en":3,"doc-seo-85858-105":30,"detail-sidebar-cat-0-en-105":92},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},85858,1099514068035,"Ezra","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","Consensus vs. Dissent Dynamic LLM Modeling of Subjective Preferences in Group Recommenders","Group recommender systems depend on how preferences are distributed within a group, yet existing social choice aggregation methods are static and cannot be connected to subjective satisfaction, fairness, and consensus judgments in real time. This paper fine-tunes large language models on human survey data to act as judgmental evaluators, generating and ranking candidate recommendations via social choice-based aggregation. A user study (N=284) shows improved satisfaction and consensus, and the strongest fairness, satisfaction, and consensus alignment when LLM judgments incorporate interaction effects with group configuration.","Consensus vs. Dissent: Dynamic LLM Modeling of Subjective Preferences in Group Recommenders  \nCedric Waterschoot  \nMaastricht University Maastricht, The Netherlands cedric.waterschoot @[maastrichtuniversity.nl](maastrichtuniversity.nl)  \nNava Tintarev  \nMaastricht University Maastricht, The Netherlands n.tintarev @[maastrichtuniversity.nl](maastrichtuniversity.nl)  \nFrancesco Barile  \nMaastricht University Maastricht, The Netherlands f.barile @[maastrichtuniversity.nl](maastrichtuniversity.nl)  \narXiv :2607 . 10235v 1 [ cs .CL] 11 Jul 2026  \nAbstract  \nPrevious work in group recommender systems has demonstrated a sensitivity to the distribution of preferences within a group. Specifically, the selection of the preference aggregation strategy benefits from considering such group configurations. In this paper, we study whether LLMs are able to mimic this sensitivity and to select the ideal aggregation strategy (and corresponding recommendation) according to nuanced human perceptions of fairness, satisfaction, and consensus.  \nWe do this by fine-tuning Large Language Models (LLMs) on human survey data to serve as real-time judgmental models within the recommendation pipeline. Using a reasoning dataset distilled from DeepSeek-V3.1 and human ground truth assessments, we develop Judgmental Llama and Judgmental OLMo to simulate group assessments. Our pipeline successfully generates multiple recommendation candidates based on social choice-based aggregation strategies and dynamically selects the one that maximizes these predicted human-like evaluations. We further validate these suggestions in a user study (􀀽 = 284) and find that our methodology achieved the highest scores for satisfaction and group consensus. Furthermore, we find that LLM judgments are most aligned with human perceptions of fairness, satisfaction and consensus when we also consider interaction effects between our LLM-based method and group configuration (e.g., minority or coalition) . These findings give further support for dynamically adapting aggregation strategies to specific within-group preference distributions, and highlight the advantage of using LLMs for an adaptation that is aligned with subjective human judgments.  \nCCS Concepts  \n• Information systems → Recommender systems; • Computing methodologies → Natural language generation.  \nKeywords  \nLarge Language Models, Fairness, Group Recommender Systems, LLM-as-judge, User study  \nPermission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission [and/or a fee. Request permissions from permissions@acm.org](and/or a fee. Request permissions from permissions@acm.org).  \nConference’17, Washington, DC, USA  \n© 2026 Copyright held by the owner/author(s) . Publication rights licensed to ACM. ACM ISBN 978-x-xxxx-xxxx-x/YYYY/MM [https://doi.org/10.1145/nnnnnnn.nnnnnnn](https://doi.org/10.1145/nnnnnnn.nnnnnnn)  \nACM Reference Format:  \nCedric Waterschoot, Nava Tintarev, and Francesco Barile. 2026. Consensus vs. Dissent: Dynamic LLM Modeling of Subjective Preferences in Group Recommenders. In . ACM, New York, NY, USA, 10 pages. [https://doi.org/10](https://doi.org/10) . 1145/nnnnnnn.nnnnnnn  \n1 Introduction  \nThe challenge for Group Recommender Systems (GRS) is the inherent difficulty of generating and adapting recommendations to the nuanced, sometimes conflicting, preferences of distinct group members. GRS research has often relied on social choice-based aggregation strategies, such as Average, Least Misery, or Most Pleasure, to bridge the gap between individual rating","cbCaifEahZG57ucH","https://ap.wps.com/l/cbCaifEahZG57ucH","pdf",879429,5,1,10,"English","en",105,"# Introduction\n## Problem: static aggregation vs. subjective group judgments\n## Proposed approach: LLM fine-tuning as judgmental models\n## Evaluation: user study and alignment with fairness, satisfaction, consensus","[{\"question\":\"What gap does the paper address in group recommender systems?\",\"answer\":\"User studies capture subjective satisfaction, fairness, and consensus, but those metrics cannot be easily integrated into automated, real-time recommendation pipelines. The paper targets the lack of a scalable mechanism to predict how a specific group will feel before recommendations are shown.\"},{\"question\":\"How do the proposed LLM-based models work in the recommendation pipeline?\",\"answer\":\"The method fine-tunes LLMs on human survey data to become judgmental models. It generates multiple recommendation candidates using social choice aggregation strategies, then uses the fine-tuned models to produce multiple assessments and selects the option that maximizes predicted human-like evaluations.\"},{\"question\":\"What were the key findings from the user study?\",\"answer\":\"With 284 participants, the dynamic LLM-based methodology achieved the highest scores for satisfaction and group consensus. The paper also finds that LLM judgments align best with perceived fairness, satisfaction, and consensus when interaction effects with group configuration are considered.\"}]",1784206745,25,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":28},"consensus-vs-dissent-dynamic-llm-modeling-of-subjective-preferences-in-group-recommenders","",{"@graph":36,"@context":86},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/consensus-vs-dissent-dynamic-llm-modeling-of-subjective-preferences-in-group-recommenders/85858/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":24,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-07-24","2026-07-16",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"What gap does the paper address in group recommender systems?","Question",{"text":76,"@type":77},"User studies capture subjective satisfaction, fairness, and consensus, but those metrics cannot be easily integrated into automated, real-time recommendation pipelines. The paper targets the lack of a scalable mechanism to predict how a specific group will feel before recommendations are shown.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How do the proposed LLM-based models work in the recommendation pipeline?",{"text":81,"@type":77},"The method fine-tunes LLMs on human survey data to become judgmental models. It generates multiple recommendation candidates using social choice aggregation strategies, then uses the fine-tuned models to produce multiple assessments and selects the option that maximizes predicted human-like evaluations.",{"name":83,"@type":74,"acceptedAnswer":84},"What were the key findings from the user study?",{"text":85,"@type":77},"With 284 participants, the dynamic LLM-based methodology achieved the highest scores for satisfaction and group consensus. The paper also finds that LLM judgments align best with perceived fairness, satisfaction, and consensus when interaction effects with group configuration are considered.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":93},[94,98,102,106,110,115,120,123,128,131,134],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":22,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":22,"slug":133},"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":20,"slug":137},19,"General","general"]