[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-579251-105":59,"doc-detail-579251-en":131},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":124,"head_meta":126,"extra_data":128,"updated_unix":130},105,"en","gaiva-a-generative-ai-framework-for-humanai-collaboration-in-educational-video-analysis","GAIVA: A Generative AI Framework for Human–AI Collaboration in Educational Video Analysis","","Qualitative video analysis is essential in learning sciences but is time-consuming and difficult to standardize, especially when ensuring consistent behavioral transcription, segmentation, and coding across researchers and contexts. GAIVA presents a theory-informed, scalable framework that combines multimodal large language models (MLLMs) and large language models (LLMs) through four sequential phases: preprocessing, MLLM transcription, human–AI codebook development, and automated coding. A simulation-based collaborative science case study shows rich transcriptions, stable theoretically grounded coding schemes, and inter-rater reliability comparable to human annotation, including the emergence of a new construct, Peripheral Engagement.",{"@graph":69,"@context":123},[70,84,106],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/gaiva-a-generative-ai-framework-for-humanai-collaboration-in-educational-video-analysis/579251/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/gaiva-a-generative-ai-framework-for-humanai-collaboration-in-educational-video-analysis/579251.png","ImageObject",300,407,{"name":92,"@type":93},"Liam","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-11-16","2026-10-15",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",14,{"@type":107,"mainEntity":108},"FAQPage",[109,115,119],{"name":110,"@type":111,"acceptedAnswer":112},"What problem does GAIVA address in educational video analysis?","Question",{"text":113,"@type":114},"GAIVA targets the labor-intensive nature of qualitative video analysis and the consistency challenges in transcription, segmentation, and coding decisions across researchers and time.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"How does GAIVA work at a high level?",{"text":118,"@type":114},"GAIVA integrates multimodal LLMs and LLMs in four phases: video preprocessing, MLLM-based behavioral transcription, human–AI collaborative codebook development grounded in CSCL theory, and LLM-based automated coding.",{"name":120,"@type":111,"acceptedAnswer":121},"What evidence does the paper provide for GAIVA’s effectiveness?",{"text":122,"@type":114},"Using a simulation-based collaborative science learning case study, GAIVA produces rich behavioral transcriptions, supports stable and theory-grounded coding schemes, and achieves inter-rater reliability comparable to human annotation.","https://schema.org",{"og:url":83,"og:type":125,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":127,"canonical":83},"index,follow",{"doc_id":129,"site_id":62},579251,1792210340,{"code":4,"msg":5,"data":132},{"doc_id":129,"user_id":133,"nickname":92,"user_avatar":134,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":135,"file_id":136,"file_url":137,"file_type":138,"file_size":139,"view_count":105,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":140,"language":141,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":142,"faqs":143,"seo_title":144,"seo_description":67,"update_tm":145,"read_time":146,"preview_page_count":4},8796095461564,"https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d","GAIVA: A Generative AI Framework for Human–AI Collaboration in Educational Video Analysis  \nYiqiu Zhou University of Illinois Urbana-Champaign Champaign, USA [yiqiuz3@illinois.edu](yiqiuz3@illinois.edu)  \nJina Kang University of Illinois Urbana-Champaign Champaign, USA [jinakang@illinois.edu](jinakang@illinois.edu)  \nHa Nguyen University of North Carolina at Chapel Hill, Chapel Hill, USA [Ha.nguyen@unc.edu](Ha.nguyen@unc.edu)  \nQualitative video analysis is a foundational yet labor-intensive method in the learning sciences, particularly for investigating the multimodal and collaborative dimensions of learning. This paper introduces GAIVA (Generative Artificial Intelligence for Video Analysis), a novel framework that integrates multimodal large language models (MLLMs) and large language models (LLMs) to support scalable, consistent, and theoryinformed video analysis. GAIVA comprises four sequential phases: (1) video preprocessing, (2) MLLMbased behavioral transcription,(3) human–AI collaborative codebook development grounded in theoretical constructs from computer-supported collaborative learning (CSCL), and (4) LLM-based automated coding. Drawing on a case study to analyze simulation-based collaborative science learning, we demonstrate that GAIVA generates rich behavioral transcriptions, enables the construction of stable and theoretically grounded coding schemes, and achieves inter-rater reliability comparable to human annotation. Notably, findings illustrate GAIVA’s capacity to support both deductive alignment and inductive discovery. The framework surfaces a behavioral construct—Peripheral Engagement—that is consistent across data batches but absent from existing frameworks. We contribute a systematic and replicable framework for integrating generative AI into qualitative video research, enabling large-scale, multimodal analysis of collaborative learning while preserving analytical rigor, theoretical coherence, and interpretive transparency.  \nKeywords: video analysis, generative AI, human–AI collaboration, multimodal large language models, large language models, collaborative learning, qualitative coding, computer-supported collaborative learning  \n1. INTRODUCTION  \nVideo analysis has long been a foundational method in learning sciences and computer-supported collaborative learning (CSCL) research for investigating the complex interplay of cognitive, social, and embodied dimensions of learning (Derry et al., 2010; Jewitt, 2012) . Unlike other forms of data (e.g., log files, discourse transcript, audio data), video preserves the rich interplay of multimodal data: verbal exchanges, embodied actions, gaze attention, and tool use  \n341 Journal of Educational Data Mining, Volume 18, No 1, 2026  \nthat characterizes authentic learning processes (Heath et al., 2010; Barron et al., 2012) . This richness has enabled researchers to examine learning in context and at multiple levels of granularity, making video a valuable tool for a wide range of studies, including investigations of teacher professional development, classroom orchestration, collaborative problem-solving, and inquiry-based learning (Powell et al., 2003; Hollingsworth, 2005; Calandra and Rich, 2014; Cukurova et al., 2020) .  \nHowever, the analytical affordance of video comes with significant methodological challenges. Qualitative video analysis is highly time-consuming and labor-intensive, often requiring dozens of hours to analyze just a single hour of footage (Markle et al., 2011; Snell, 2011; Lemke, 2014) . In addition, video analysis faces challenges with consistency. Researchers must make decisions about which behaviors to attend to, how to segment video episodes, and how to interpret actions within context. These decisions may vary across researchers or even within the same researcher over time, raising concerns about reliability and replicability (Erickson, 2006; Lemke, 2014) . The absence of standardized protocols to guide these choices can lead to reduced trans","cbCaiegUXo9AUKWT","https://ap.wps.com/l/cbCaiegUXo9AUKWT","pdf",701875,46,"English","# Introduction\n## Motivation and challenges in qualitative video analysis\n## Related work and limitations of prior tools\n## Generative AI opportunities with multimodal LLMs","[{\"question\":\"What problem does GAIVA address in educational video analysis?\",\"answer\":\"GAIVA targets the labor-intensive nature of qualitative video analysis and the consistency challenges in transcription, segmentation, and coding decisions across researchers and time.\"},{\"question\":\"How does GAIVA work at a high level?\",\"answer\":\"GAIVA integrates multimodal LLMs and LLMs in four phases: video preprocessing, MLLM-based behavioral transcription, human–AI collaborative codebook development grounded in CSCL theory, and LLM-based automated coding.\"},{\"question\":\"What evidence does the paper provide for GAIVA’s effectiveness?\",\"answer\":\"Using a simulation-based collaborative science learning case study, GAIVA produces rich behavioral transcriptions, supports stable and theory-grounded coding schemes, and achieves inter-rater reliability comparable to human annotation.\"}]","GAIVA: A Generative AI Framework for Human–AI Collaboration in Educational Video Analysis | PDF",1792035287,116]