[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82311-en":3,"doc-seo-82311-105":28,"detail-sidebar-cat-0-en-105":90},{"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":20,"is_downloadable":20,"audit_status":20,"page_count":11,"language":21,"language_code":22,"site_id":23,"html_lang":22,"table_of_contents":24,"faqs":25,"seo_title":13,"seo_description":14,"update_tm":26,"read_time":27},82311,1374391974564,"Clementine","https://ap-avatar.wpscdn.com/avatar/14000253aa45c000a9e?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779874745381141002",8,"Research & Report","Towards Detecting Inconsistencies in End-to-end Generated TODs","Generative AI is reshaping conversational systems toward end-to-end task-oriented dialogue (TOD) generation, yet large language models can still produce inconsistencies that break task success. A single hallucination—such as recommending a nonexistent restaurant—can cause severe failures because TOD responses must strictly follow a domain knowledge base. The work formulates TOD generation as a constraint satisfaction problem (CSP), detects inconsistent dialogue segments, and proposes minimal edits to restore consistency, reporting high accuracy and an in-depth analysis of results.","Towards Detecting Inconsistencies in End-to-end Generated TODs  \nTiziano Labruna  \n[tlabruna@fbk.eu](tlabruna@fbk.eu)[ ](tlabruna@fbk.eu)Fondazione Bruno Kessler Povo, Trento, Italy  \nGiovanni Bonetta  \n[gbonetta@fbk.eu](gbonetta@fbk.eu)[ ](gbonetta@fbk.eu)Fondazione Bruno Kessler Povo, Trento, Italy  \nBernardo Magnini  \n[magnini@fbk.eu](magnini@fbk.eu)[ ](magnini@fbk.eu)Fondazione Bruno Kessler Povo, Trento, Italy  \narXiv :2607 .09338v 1 [ cs .CL] 10 Jul 2026  \nAbstract  \nGenerative AI is profoundly transforming the core technologies behind conversational systems, shifting from component-based to end-to-end approaches. However, Large Language Models (LLMs) may still generate inconsistencies, a critical issue particularly in Task-Oriented Dialogues (TODs), where system responses must strictly adhere to information from a domain knowledge base (e.g., restaurants in a city). A single hallucination (e.g., suggesting a nonexistent restaurant) can lead to severe task failures. We investigate a method for automatically detecting inconsistencies by conceptualizing TODs as a Constraint Satisfaction Problem (CSP), where variables represent dialogue segments referencing the conversational domain, and constraints among variables capture dialogue properties such as turn coherence and adherence to domain knowledge. We propose a pipeline that first identifies variables in a target dialogue and then applies a CSP solver to identify valid solutions. By comparing the target dialogue with valid variable assignments, we can detect inconsistencies and suggest minimal changes to ensure dialogue consistency. We demonstrate the high accuracy of the CSP-based approach in detecting inconsistencies, and provide a detailed analysis of our findings.  \nCCS Concepts  \n• Computing methodologies → Discourse, dialogue and pragmatics; Natural language generation; Language resources.  \nKeywords  \nTask-Oriented Dialogue Systems, Dialogue Consistency, Constraint Satisfaction Problem, Large Language Models.  \nACM Reference Format:  \nTiziano Labruna, Giovanni Bonetta, and Bernardo Magnini. 2026. Towards Detecting Inconsistencies in End-to-end Generated TODs. In Proceedings of Proceedings of the SIGIR Workshop on Search-Oriented Conversational AI (SCAI). ACM, New York, NY, USA, 8 pages. [https://doi.org/10.1145/nnnnnnn](https://doi.org/10.1145/nnnnnnn). nnnnnnn  \n1 Introduction  \nTask-oriented dialogue (TOD) systems [3, 6, 24, 28] play a crucial role in human-computer interaction, facilitating seamless communication between users and machines to perform specific tasks. In  \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).  \nSCAI, Melbourne, Australia  \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)  \nKnowledge Base  \n\n| ID | Name | Area | Food | Price |\n| --- | --- | --- | --- | --- |\n| R1 | Taberna | centre | spanish | cheap |\n| R2 | Espana | centre | spanish | moderate |\n| R3 | Beirut | centre | lebanese | cheap |\n\nDialogue  \nUser: I am looking for a restaurant serving Spanish  \nfood.  \nSystem: There are three restaurants serving Spanish food, one is cheap and the other is moderate price range. Which price range would you prefer?  \nUser: I am looking for a cheap restaurant in any area that serves Spanish food.  \nSystem: Beiru","cbCaih8oclOTmy3y","https://ap.wps.com/l/cbCaih8oclOTmy3y","pdf",647124,1,"English","en",105,"# Introduction\n## Knowledge base and inconsistent TOD example\n# Method: CSP-based inconsistency detection\n## Variable identification and constraint modeling\n## CSP solving and minimal change suggestions\n# Results and analysis\n## Accuracy of the CSP approach\n# Conclusion","[{\"question\":\"Why do inconsistencies matter in end-to-end generated task-oriented dialogues (TODs)?\",\"answer\":\"TOD systems must align system responses with a domain knowledge base. Even one hallucinated detail (e.g., a nonexistent restaurant) can make the dialogue fail the task.\"},{\"question\":\"How does the proposed method detect inconsistencies?\",\"answer\":\"It conceptualizes TODs as a constraint satisfaction problem (CSP), where dialogue segments become variables and constraints capture turn coherence and adherence to domain knowledge. A CSP solver finds valid assignments, which are compared against the target dialogue.\"},{\"question\":\"What kinds of inconsistencies does the document distinguish, and how are they illustrated?\",\"answer\":\"It distinguishes domain inconsistencies (slot values that contradict the knowledge base) from dialogic inconsistencies (content that is inconsistent with prior dialogue turns). An example dialogue with a restaurant KB shows both types, tied to specific turns.\"}]",1784179537,20,{"code":4,"msg":29,"data":30},"ok",{"site_id":23,"language":22,"slug":31,"title":13,"keywords":32,"description":14,"schema_data":33,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":26},"towards-detecting-inconsistencies-in-end-to-end-generated-tods","",{"@graph":34,"@context":84},[35,52,67],{"@type":36,"itemListElement":37},"BreadcrumbList",[38,42,46,49],{"item":39,"name":40,"@type":41,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":43,"name":44,"@type":41,"position":45},"https://docshare.wps.com/document/","Document",2,{"item":47,"name":12,"@type":41,"position":48},"https://docshare.wps.com/document/research-report/",3,{"item":50,"name":13,"@type":41,"position":51},"https://docshare.wps.com/document/towards-detecting-inconsistencies-in-end-to-end-generated-tods/82311/",4,{"url":50,"name":13,"@type":53,"author":54,"headline":13,"publisher":56,"fileFormat":59,"inLanguage":22,"description":14,"dateModified":60,"datePublished":61,"encodingFormat":59,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":55},"Person",{"url":39,"name":57,"@type":58},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":20},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"Why do inconsistencies matter in end-to-end generated task-oriented dialogues (TODs)?","Question",{"text":74,"@type":75},"TOD systems must align system responses with a domain knowledge base. Even one hallucinated detail (e.g., a nonexistent restaurant) can make the dialogue fail the task.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"How does the proposed method detect inconsistencies?",{"text":79,"@type":75},"It conceptualizes TODs as a constraint satisfaction problem (CSP), where dialogue segments become variables and constraints capture turn coherence and adherence to domain knowledge. A CSP solver finds valid assignments, which are compared against the target dialogue.",{"name":81,"@type":72,"acceptedAnswer":82},"What kinds of inconsistencies does the document distinguish, and how are they illustrated?",{"text":83,"@type":75},"It distinguishes domain inconsistencies (slot values that contradict the knowledge base) from dialogic inconsistencies (content that is inconsistent with prior dialogue turns). An example dialogue with a restaurant KB shows both types, tied to specific turns.","https://schema.org",{"og:url":50,"og:type":86,"og:title":13,"og:site_name":57,"og:description":14},"article",{"robots":88,"canonical":50},"index,follow",{"doc_id":7,"site_id":23},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,122,126,129,133],{"id":20,"doc_module":4,"doc_module_name":44,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":45,"doc_module":4,"doc_module_name":44,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":51,"doc_module":4,"doc_module_name":44,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":44,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":44,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":44,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":44,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":44,"category_name":124,"show_sort_weight":27,"slug":125},9,"Religion & Spirituality","religion-spirituality",{"id":27,"doc_module":4,"doc_module_name":44,"category_name":127,"show_sort_weight":27,"slug":128},"World Cup","world-cup",{"id":130,"doc_module":4,"doc_module_name":44,"category_name":131,"show_sort_weight":130,"slug":132},10,"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":44,"category_name":135,"show_sort_weight":105,"slug":136},19,"General","general"]