[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85610-en":3,"doc-seo-85610-105":29,"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":20,"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":13,"seo_description":14,"update_tm":27,"read_time":28},85610,3848291630094,"Emma Wilson","https://eur-avatar.wpscdn.com/davatar_085a072bc5b1113ac321206ff7593b45",8,"Research & Report","Severity-Aware Curriculum Learning with Multi-Model Response Selection for Medical Text Generation","Telehealth increasingly depends on medical language generation, yet large language models can produce inconsistent and context-inappropriate answers when case severity varies. This work proposes a severity-aware multi-model framework that couples a three-stage curriculum learning strategy with relevance-based response selection. Five large language models are trained sequentially on mild, moderate, and critical cases under a shared curriculum, then generate candidates at inference. The response with the highest BERTScore is selected. Experiments on the MAQA dataset achieve 90.30% after fine-tuning, improving quality and relevance.","Severity-Aware Curriculum Learning with Multi-Model Response Selection for Medical Text  \nGeneration  \nAhmed Alansary  \nFaculty of Computer Science MSA University Giza, Egypt  \n[ahmed.mohamed406@msa.edu.eg](ahmed.mohamed406@msa.edu.eg)  \nMolham Mohamed  \nFaculty of Computer Science MSA University Giza, Egypt  \n[molham.mohamed@msa.edu.eg](molham.mohamed@msa.edu.eg)  \nAli Hamdi  \nFaculty of Computer Science MSA University Giza, Egypt [ahamdi@msa.edu.eg](ahamdi@msa.edu.eg)  \narXiv :2606 .055 10v 3 [ cs .AI] 13 Jul 2026  \nAbstract—Telehealth systems have become increasingly important for delivering accessible and timely medical information. Existing large language models often struggle to provide consistent and contextually appropriate medical responses across varying levels of case severity. This limitation highlights the need for models that can effectively adapt to the progressive complexity in medical queries. To address this challenge, we introduce a severity-aware multi-model framework that integrates curriculum training strategy with relevance-based response selection. The proposed framework employs a three-stage curriculum learning strategy, where each model is trained sequentially on mild, moderate, and critical cases to progressively acquire domain knowledge. The approach uses five large language models, each trained independently under the same curriculum. During inference, all models generate candidate responses, and the response with highest BERTScore is selected as the final output. The framework is trained and evaluated on the MAQA dataset, which provides annotated medical questionanswer pairs. Experimental results evaluated using BERTScore demonstrate that the proposed method achieves superior performance compared to both baseline and fine-tuned models, attaining 86.71% in the baseline setting and 90.30% after fine-tuning. These results highlight the effectiveness of combining curriculum learning with multi-model response selection in improving response quality and relevance in medical text generation.  \nIndex Terms—Curriculum Learning, Arabic Medical Text Generation, Large Language Models, Natural Language Processing, Multi-Model Response Selection, Severity-aware Framework.  \nI. Introduction  \nRecent progress in telehealth infrastructure has reshaped how patients interact with healthcare providers, allowing remote and timely access to medical consultations [1] . Online healthcare platforms enable users to describe symptoms and obtain medical recommendations, producing large collections of health-related textual data that can support intelligent decision-making systems [2], [3] . In  \n979-8-3315-8488-7/26/$31.00 ©2026 IEEE  \nparallel, natural language processing (NLP) plays a key role in building intelligent systems for clinical support, facilitating the understanding and generation of medical text across various applications, including clinical support, personal care, and public health [4], [5] . The growing use of medical bots highlights the importance of generating responses that are both relevant and dependable in realworld healthcare settings. [6]–[8] .  \nDespite these advances, existing approaches have several limitations. Recent progress in data-to-text generation has largely relied on neural end-to-end systems, where curriculum learning has been shown to improve both performance and speed by organizing training samples based on difficulty [9] . Similarly, in-sample curriculum strategies that follow an easy-to-hard learning paradigm have demonstrated strong generalization capabilities in natural language generation tasks [10] . In the context of healthcare applications, transformer-based architectures have shown competitive results across a range of clinical tasks including symptom analysis and disease categorization, symptom severity assessment, and medical text generation [2], [3], [11] . Ensemble learning techniques improve prediction reliability by combining outputs from multiple models [5] . Nevertheles","cbCaicPsN3ETCh3n","https://ap.wps.com/l/cbCaicPsN3ETCh3n","pdf",258795,1,6,"English","en",105,"# Abstract\n# I. Introduction\n# II. Related Work","[{\"question\":\"What is the main goal of the proposed framework?\",\"answer\":\"To improve the consistency and relevance of medical text generation by adapting responses to different case severities.\"},{\"question\":\"How does the curriculum learning strategy work in this method?\",\"answer\":\"Models are trained sequentially on mild, moderate, and critical cases, progressively increasing the complexity of medical scenarios.\"},{\"question\":\"How are final responses selected during inference?\",\"answer\":\"All models generate candidate answers, and the system selects the response with the highest BERTScore.\"}]",1784204906,15,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"severity-aware-curriculum-learning-with-multi-model-response-selection-for-medical-text-generation","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/severity-aware-curriculum-learning-with-multi-model-response-selection-for-medical-text-generation/85610/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is the main goal of the proposed framework?","Question",{"text":75,"@type":76},"To improve the consistency and relevance of medical text generation by adapting responses to different case severities.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the curriculum learning strategy work in this method?",{"text":80,"@type":76},"Models are trained sequentially on mild, moderate, and critical cases, progressively increasing the complexity of medical scenarios.",{"name":82,"@type":73,"acceptedAnswer":83},"How are final responses selected during inference?",{"text":84,"@type":76},"All models generate candidate answers, and the system selects the response with the highest BERTScore.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,114,119,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":21,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]