[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-292257-105":59,"doc-detail-292257-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","edupulse-a-practical-llm-enhanced-opinion-mining-system-for-vietnamese-student-feedback-in-educational-platforms","EduPulse - A Practical LLM-Enhanced Opinion Mining System for Vietnamese Student Feedback in Educational Platforms","","Opinion mining from real-world student feedback presents significant practical challenges, including linguistic noise such as slang and teencode, along with scalability and maintainability concerns that are often overlooked in academic work. The paper proposes EduPulse, a Vietnamese-focused opinion mining system implementing four tasks: sentiment classification, category-based sentiment, suggestion detection, and opinion summarization. A hybrid architecture combines LLM robustness for complex noisy cases with a lightweight neural model for high-throughput, low-cost operation.",{"@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/edupulse-a-practical-llm-enhanced-opinion-mining-system-for-vietnamese-student-feedback-in-educational-platforms/292257/",{"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/edupulse-a-practical-llm-enhanced-opinion-mining-system-for-vietnamese-student-feedback-in-educational-platforms/292257.png","ImageObject",300,407,{"name":92,"@type":93},"McQueen","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-29","2026-09-17",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",11,{"@type":107,"mainEntity":108},"FAQPage",[109,115,119],{"name":110,"@type":111,"acceptedAnswer":112},"What are the four main tasks implemented by EduPulse?","Question",{"text":113,"@type":114},"EduPulse performs sentiment classification, category-based sentiment classification, suggestion detection, and opinion summarization.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"How does EduPulse handle Vietnamese linguistic noise such as slang and teencode?",{"text":118,"@type":114},"It uses a hybrid architecture where Large Language Models are leveraged for robustness on complex, noise-sensitive tasks like sentiment classification and suggestion detection.",{"name":120,"@type":111,"acceptedAnswer":121},"What design goals does the hybrid architecture balance?",{"text":122,"@type":114},"The architecture strategically balances performance, cost, and maintainability, aiming to avoid expensive retraining cycles while improving task accuracy.","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},292257,1790007038,{"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},5909890329169,"https://ap-avatar.wpscdn.com/davatar_9964176cb1d06d4a9deccf72a44ae3dc","EduPulse: A Practical LLM-Enhanced Opinion Mining System for Vietnamese Student Feedback in Educational Platforms  \nNguyen Xuan Phuc* , Nguyen Xuan Phi* , Nguyen Vinh Tiep,  \nDang Van Thin, Ngan Luu-Thuy Nguyen†  \nUniversity of Information Technology-VNUHCM, Vietnam National University, Ho Chi Minh City, Vietnam  \n[23521213@gm.uit.edu.vn](23521213@gm.uit.edu.vn), [xphi.work@gmail.com](xphi.work@gmail.com)  \n{thindv,tiepnv, [ngannlt}@uit.edu.vn](ngannlt}@uit.edu.vn)  \nAbstract  \nOpinion mining from real-world student feedback presents significant practical challenges, such as handling linguistic noise (slang, teencode) and the need for scalable and maintainable systems, which are often overlooked in academic research. This paper introduces EduPulse, a practical opinion mining system designed specifically to analyze student feedback in Vietnamese. Our application 1 performs four opinion analysis tasks, including Sentiment Classification, Category-based Sentiment Classification, Suggestion Detection, and Opinion Summarization. We design the hybrid architecture that strategically balances performance, cost, and maintainability. This architecture leverages the robustness of Large Language Models (LLMs) for complex, noisesensitive tasks as sentiment classification and suggestion detection, while employing a specialized, lightweight neural model for highthroughput, low-cost solutions. Our experiments show that applying the LLM-based approach achieves high robustness, justifying its operational cost by eliminating expensive retraining cycles. Furthermore, we demonstrate that our collaborative modular architecture significantly improves task performance (+7.6%) compared to traditional approaches, offering a practical design for industry-focused Natural Language Processing applications.  \n1 Introduction  \nOpinion Mining (OM) or Sentiment Analysis (SA) is one of the most widely used NLP applications for identifying the emotions, intentions, and attitudes based on user comments or feedbacks (Liu, 2022) . Several recent studies (Wankhade et al., 2022 ; Mao et al., 2024 ; Sharma et al., 2025) comprehensively surveyed the methods, challenges,  \n*These authors contributed equally.  \n†Corresponding author: [ngannlt@uit.edu.vn](ngannlt@uit.edu.vn)  \n1A video demonstration is available at [https://www](https://www). [youtube.com/watch?v=tiWkpK-aWoI](youtube.com/watch?v=tiWkpK-aWoI)  \nand applications of OM in mining the valuable information generated from these user comments. However, most of these works primarily focus on researching methodologies to address various technical problems within the SA field, rather than detailing their practical, real-world applications. For instance, the work by Zhang et al. (2024) presented a comprehensive evaluation investigating the capabilities of Large Language Models (LLMs) across diverse sentiment analysis tasks, ranging from simple to complex. Similarly, Zhang et al. (2025b) also explored the effectiveness of larger LLMs in addressing the challenge of labeled data scarcity in the software engineering domain. In addition, several attempts (Hellwig et al., 2025 ; Xu et al., 2025b) have proposed data augmentation based on the power of LLMs for various sentiment tasks.  \nIn the education domain, mining the information embedded in student comments plays a pivotal role, particularly for comprehensive evaluation of management systems, teaching efficacy, and academic curricula (Elfeky et al., 2020) . Most universities and institutions rely heavily on student feedback for quality assurance (Shaik et al., 2023a) . However, the unstructured format, noise, and scalability of the student’s feedback often result in manual analysis becoming an operational bottleneck. To tackle this challenge, applying artificial intelligence (AI) solutions is essential for automating the processing, extraction of insights, and comprehensive evaluation of large-scale educational feedback datasets.  \nFor low-resource languages like Vietnamese, the","cbCaivsgeWnB6b5b","https://ap.wps.com/l/cbCaivsgeWnB6b5b","pdf",2644532,28,"English","# Introduction\n## Opinion Mining in Education\n## Challenges for Low-Resource Languages\n## EduPulse System and Tasks","[{\"question\":\"What are the four main tasks implemented by EduPulse?\",\"answer\":\"EduPulse performs sentiment classification, category-based sentiment classification, suggestion detection, and opinion summarization.\"},{\"question\":\"How does EduPulse handle Vietnamese linguistic noise such as slang and teencode?\",\"answer\":\"It uses a hybrid architecture where Large Language Models are leveraged for robustness on complex, noise-sensitive tasks like sentiment classification and suggestion detection.\"},{\"question\":\"What design goals does the hybrid architecture balance?\",\"answer\":\"The architecture strategically balances performance, cost, and maintainability, aiming to avoid expensive retraining cycles while improving task accuracy.\"}]","EduPulse - A Practical LLM-Enhanced Opinion Mining System for Vietnamese Student Feedback in Educational Platforms | PDF",1789649413,71]