[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-639822-105":3,"detail-sidebar-cat-0-en-105":81,"doc-detail-639822-en":130},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":74,"head_meta":76,"extra_data":78,"updated_unix":80},105,"en","learning-action-conditions-from-instructional-manuals-for-instruction-understanding","Learning Action Conditions from Instructional Manuals for Instruction Understanding","","Inferring an action’s preconditions and postconditions is critical for understanding complex instructions and for enabling autonomous, instruction-guided agents as well as assistive AI that helps humans perform physical tasks. This work introduces action condition inference, extracting mentions of action preconditions and postconditions from instructional manuals. A weakly supervised method is proposed with automatically built large-scale training instances from online instructions, alongside a densely human-annotated and validated dataset to evaluate current NLP models, showing significant F1 gains from full context and task-specific heuristics, while still lagging behind human performance.",{"@graph":14,"@context":73},[15,34,56],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/document/","Document",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/document/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/learning-action-conditions-from-instructional-manuals-for-instruction-understanding/639822/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/learning-action-conditions-from-instructional-manuals-for-instruction-understanding/639822.png","ImageObject",300,407,{"name":42,"@type":43},"jay","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-30","2026-10-22",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",7,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69],{"name":60,"@type":61,"acceptedAnswer":62},"What is action condition inference in instructional manuals?","Question",{"text":63,"@type":64},"It is a task that extracts mentions of both preconditions and postconditions of actions from instructional manuals.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"Why are preconditions and postconditions important for instruction understanding?",{"text":68,"@type":64},"They provide a structured dependency view of a task, helping agents decide whether to proceed and evaluate whether actions are completed correctly.",{"name":70,"@type":61,"acceptedAnswer":71},"How does the paper build training data for the proposed task?",{"text":72,"@type":64},"It uses a weakly supervised approach with automatically constructed large-scale training instances from online instructions, plus a densely human-annotated and validated dataset for study and evaluation.","https://schema.org",{"og:url":32,"og:type":75,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":77,"canonical":32},"index,follow",{"doc_id":79,"site_id":7},639822,1793052821,{"code":4,"msg":82,"data":83},"success",[84,88,92,96,101,106,110,114,119,122,126],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":25,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":102,"doc_module":4,"doc_module_name":25,"category_name":103,"show_sort_weight":104,"slug":105},6,"Technology",50,"technology",{"id":55,"doc_module":4,"doc_module_name":25,"category_name":107,"show_sort_weight":108,"slug":109},"Healthcare",40,"healthcare",{"id":111,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":112,"slug":113},8,30,"research-report",{"id":115,"doc_module":4,"doc_module_name":25,"category_name":116,"show_sort_weight":117,"slug":118},9,"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":25,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":25,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":25,"category_name":128,"show_sort_weight":97,"slug":129},19,"General","general",{"code":4,"msg":82,"data":131},{"doc_id":79,"user_id":132,"nickname":42,"user_avatar":133,"doc_module":4,"category_id":111,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":139,"language":140,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":12,"update_tm":144,"read_time":145},3985747858866,"https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd","Learning Action Conditions from Instructional Manuals for Instruction Understanding  \nTe-Lin Wu 1 , Caiqi Zhang2 , Qingyuan Hu 1 , Alex Spangher3 , Nanyun Peng 1  \n1University of California, Los Angeles, 2University of Cambridge, 3Information Sciences Institute, University of Southern California  \n{telinwu,violetpeng,[hu528}@cs.ucla.edu](hu528}@cs.ucla.edu), [cz391@cam.ac.uk](cz391@cam.ac.uk), [spangher@isi.edu](spangher@isi.edu)  \nAbstract  \nThe ability to infer pre-and postconditions of an action is vital for comprehending complex instructions, and is essential for applications such as autonomous instruction-guided agents and assistive AI that supports humans to perform physical tasks. In this work, we propose atask dubbed action condition inference, which extracts mentions of preconditions and postconditions of actions in instructional manuals. We propose a weakly supervised approach utilizing automatically constructed large-scale training instances from online instructions, and curate a densely human-annotated and validated dataset to study how well the current NLP models do on the proposed task. We design two types of models differ by whether contextualized and global information is leveraged, as well as various combinations of heuristics to construct the weak supervisions. Our experiments show a >20% F1-score improvement with considering the entire instruction contextsand a > 6% F1-score benefit with the proposed heuristics. However, the best performing model is still well-behind human performance.1  \n1 Introduction  \nWhen performing complex tasks (e.g. making a gourmet dish), instructional manuals are often referred to as useful guidelines. To follow the instructed actions, it is crucial to understand the preconditions, i.e. prerequisites before taking a particular action, and the postconditions, i.e. the status supposed to be reached after performing the action. Knowledge of action-condition dependencies is prevalent and inferable in many instructional texts. For example, in Figure 1, before performing the action “place onions\" in step 3, both preconditions: “heat the pan\" (in step 2) and “slice onions\"(in step 1) have to be successfully accomplished. Likewise, executing “stir onions\" (in step 4), leads to its postcondition,“caramelized\" (also in step 4) .  \n1Dataset and codes will be released at: here.  \nFigure 1: The Action Condition Inference Task: We propose a task that probes models’ ability to infer both preconditions and postconditions of an action from instructional manuals. It has wide applications to e.g. assistive AI and tasksolving robots. ∗This instruction is simplified for illustration.  \nFor autonomous agents or assistant AI that aids humans to accomplish tasks, understanding the conditions provides a structured view of a task (Linden, 1994 ; Aeronautiques et al., 1998 ; Branavan et al., 2012a ; Sharma and Kroemer, 2020) and helps the agent correctly judge whether to proceed to the next action and evaluate the action completions. However, no prior work has systematically studied automatically extracting pre-and postconditions from prevalent data resources. To bridge this gap, we propose the action condition inference task on real-world instructional manuals, where a dense dependency graph is produced, as in Figure 1, to denote the pre-and postconditions of actions. Such a dependency graph provides a systematic task execution plan that agents can closely follow. We consider two online instruction resources, WikiHow (Hadley et al.) [and](and Instructables.com)[ Instructables.com](and Instructables.com) (Instructables), to study the current NLP models’ capabilities of performing the proposed task. As there is no densely annotated dataset on the desired action-condition-dependencies from real-world instructions, and annotating a comprehensive depen-  \n3023  \nProceedings of the 61st Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers, pages 3023–3043 July 9-14, 2023 ©2023 Ass","cbCaiehbtkwA9tfL","https://ap.wps.com/l/cbCaiehbtkwA9tfL","pdf",1982269,21,"English","# Abstract\n# Introduction\n# Dataset and codes\n# Figure 1: The Action Condition Inference Task\n# Figure 2: Terminologies\n# SRL Sample Extractions","[{\"question\":\"What is action condition inference in instructional manuals?\",\"answer\":\"It is a task that extracts mentions of both preconditions and postconditions of actions from instructional manuals.\"},{\"question\":\"Why are preconditions and postconditions important for instruction understanding?\",\"answer\":\"They provide a structured dependency view of a task, helping agents decide whether to proceed and evaluate whether actions are completed correctly.\"},{\"question\":\"How does the paper build training data for the proposed task?\",\"answer\":\"It uses a weakly supervised approach with automatically constructed large-scale training instances from online instructions, plus a densely human-annotated and validated dataset for study and evaluation.\"}]","Learning Action Conditions from Instructional Manuals for Instruction Understanding | PDF",1792705587,53]