[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-123628-en":3,"doc-seo-123628-105":30,"detail-sidebar-cat-0-en-105":91},{"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":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},123628,687197100911,"Himbo","https://ap-avatar.wpscdn.com/avatar/a000239b6f1da00475?x-image-process=image/resize,m_fixed,w_180,h_180&k=1785132997149421697",8,"Research & Report","Addressing Contingency in Algorithmic (Mis)information Classification - Toward a Responsible Machine Learning Agenda","Machine learning classification models are increasingly used to detect and manage online misinformation, raising the need to evaluate how “truth” sources are treated as legitimate, authoritative, and objective during model training and testing. The political, ethical, and epistemic stakes are often omitted in technical work, even though ML moderation can reshape public debate and produce harms like undue censorship and reinforced false beliefs. Using collaborative ethnography, the study analyzes algorithmic contingencies and proposes a reflexive, responsible development agenda for moderation systems.","Addressing contingency in algorithmic (mis)information classification: Toward a responsible machine learning agenda  \nAndrés Domínguez Hernández*  \nDepartment of Computer Science, University of Bristol, [andres.dominguez@bristol.ac.uk](andres.dominguez@bristol.ac.uk)  \nRichard Owen  \nSchool of Management, University of Bristol  \nDan Saattrup Nielsen  \nDepartment of Engineering Mathematics, University of Bristol  \nRyan McConville  \nDepartment of Engineering Mathematics, University of Bristol  \n*Corresponding author  \nAbstract  \nMachine learning (ML) enabled classification models are becoming increasingly popular for tackling the sheer volume and speed of online misinformation and other content that could be identified as harmful. In building these models, data scientists need to take a stance on the legitimacy, authoritativeness and objectivity of the sources of “truth” used for model training and testing. This has political, ethical and epistemic implications which are rarely addressed in technical papers. Despite (and due to) their reported high accuracy and performance, ML-driven moderation systems have the potential to shape online public debate and create downstream negative impacts such as undue censorship and the reinforcing of false beliefs. Using collaborative ethnography and theoretical insights from social studies of science and expertise, we offer a critical analysis of the process of building ML models for (mis)information classification: we identify a series of algorithmic contingencies—key moments during model development that could lead to different future outcomes, uncertainty and harmful effects as these tools are deployed by social media platforms. We conclude by offering a tentative path toward reflexive and responsible development of ML tools for moderating misinformation and other harmful content online.  \nCCS Concepts: • Information systems~Information systems applications~Collaborative and social computing systems and tools~Social networking sites • General and reference~Cross-computing tools and techniques~Evaluation  \nKeywords and Phrases: misinformation, content moderation, fact-checking, machine learning, responsible ML  \nACM Reference Format:  \nAndrés Domínguez Hernández, Richard Owen, Dan Saattrup Nielsen and Ryan McConville. 2023. Addressing contingency in algorithmic (mis)information classification: Toward a responsible machine learning agenda. In 2023 ACM Conference on Fairness, Accountability, and Transparency (FAccT ’23), June 12–15, 2023, Chicago, United States of America. ACM, New York, NY, USA, 16 pages.  \n1 Introduction  \nIn recent years there has been a flurry of research on the automated detection of misinformation using Machine Learning (ML) techniques. Significant progress has been made on developing ML models for the identification, early detection and management of online misinformation, 1 which can then be deployed at scale to assist human moderators, e.g.,[34,50,77] . The development of these tools has gained currency particularly among social media platforms like Meta, Twitter and YouTube given their key role in the propagation of online misinformation and mounting regulatory pressure to manage the problem. In response to the overwhelming scale of misinformation –notably in the context of the COVID-19 pandemicand the limited capacity of human moderation to address this, platforms have increasingly looked to the deployment of automated models as standalone solutions requiring less or no human intervention [9] .  \nThe artificial intelligence (AI) research community has broadly framed the problem as one that can be tackled using ML–enabled classification models. These classify, with varying levels of accuracy, the category to which a piece of data belongs (e.g. “factually true”, “false” or “misleading” claim) . These models are trained on large datasets of various modalities (images, text or social connections) containing manually annotated samples of information labelled as being fa","cbCaipde6dBdvh9E","https://ap.wps.com/l/cbCaipde6dBdvh9E","pdf",499476,1,17,"English","en",105,"# Introduction\n## Algorithmic framing of misinformation detection\n## Training data, “ground truths,” and epistemic assumptions\n## Algorithmic contingencies and omitted ethical implications","[{\"question\":\"Why do ML systems used for misinformation classification require attention to the concept of “truth” sources?\",\"answer\":\"Model training and evaluation depend on sources treated as legitimate and authoritative ground truths. The study argues that these epistemic assumptions carry political, ethical, and social implications that are rarely addressed in technical papers.\"},{\"question\":\"What harms can arise from deploying ML-driven moderation models despite high reported accuracy?\",\"answer\":\"The document highlights risks such as undue censorship and reinforcing false beliefs, since moderation systems can influence online public debate and downstream outcomes.\"},{\"question\":\"What are “algorithmic contingencies” in the proposed analysis?\",\"answer\":\"Algorithmic contingencies are key moments during model development that can lead to different future outcomes, producing uncertainty and harmful effects when such tools are deployed by social media platforms.\"}]","Addressing Contingency in Algorithmic (Mis)information Classification - Toward a Responsible Machine Learning Agenda | PDF",1785817714,43,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"addressing-contingency-in-algorithmic-misinformation-classification-toward-a-responsible-machine-learning-agenda","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"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/addressing-contingency-in-algorithmic-misinformation-classification-toward-a-responsible-machine-learning-agenda/123628/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-04",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why do ML systems used for misinformation classification require attention to the concept of “truth” sources?","Question",{"text":75,"@type":76},"Model training and evaluation depend on sources treated as legitimate and authoritative ground truths. The study argues that these epistemic assumptions carry political, ethical, and social implications that are rarely addressed in technical papers.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What harms can arise from deploying ML-driven moderation models despite high reported accuracy?",{"text":80,"@type":76},"The document highlights risks such as undue censorship and reinforcing false beliefs, since moderation systems can influence online public debate and downstream outcomes.",{"name":82,"@type":73,"acceptedAnswer":83},"What are “algorithmic contingencies” in the proposed analysis?",{"text":84,"@type":76},"Algorithmic contingencies are key moments during model development that can lead to different future outcomes, producing uncertainty and harmful effects when such tools are deployed by social media platforms.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"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":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]