Scholarlink

<p>Autism spectrum disorder (ASD) affects tens of millions of families worldwide, yet parents confront abundant but unreliable online advice and limited access to timely, empathetic guidance. To address this critical gap, we developed Starmate (<a href="http://kefeng.mpu.edu.mo/starmate">http://kefeng.mpu.edu.mo/starmate</a>), a 1.5B-parameter, domain-tuned AI assistant for ASD caregivers, using a rigorous user-centered mixed-methods framework. Informed by in-depth interviews (<InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(n=13\)</EquationSource> </InlineEquation>) and a Kano survey (<InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(n=60\)</EquationSource> </InlineEquation>) that identified “Hands-on guidance” as a must-have caregiver requirement, we engineered a novel modular architecture that integrates sentiment analysis, expert-vetted knowledge-graph-augmented retrieval (LightRAG), and a domain-fine-tuned Qwen2.5-1.5B model. In a blinded, side-by-side comparison against leading commercial LLMs, Starmate demonstrated improved performance across key metrics within this evaluation framework (86.76 vs 78.43–83.84; <InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(p &lt; 0.001\)</EquationSource> </InlineEquation>) and showed specific advantages in Empathy, Hands-on guidance, and Logical clarity (all <InlineEquation ID="IEq4"> <EquationSource Format="TEX">\(p &lt; 0.05\)</EquationSource> </InlineEquation>). Automated benchmarking corroborated these results, with top scores for Professional accuracy (86.18), Empathy (86.79), and Hands-on guidance (82.58). These findings demonstrate the technical feasibility of a lightweight, privacy-conscious, domain-specific LLM to generate accurate, empathetic, and actionable responses in benchmarked scenarios, laying the groundwork for future real-world usability and clinical testing.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Starmate: A Lightweight AI Assistant for Autism Caregivers Developed and Evaluated Through a User-Centered Mixed-Methods Framework

  • 期刊论文

  • 发表日期：

    2026年7月2日

查看全文
  • Zhifan Li,
  • Xiaoxia Liu,
  • Tianhao Chen,
  • Yuting Yang,
  • Xiaoyan Liu,
  • Yuanyuan Lv,
  • Zixuan Zhao,
  • Xueying Li,
  • Xiaoqing Yin,
  • Zhongwen Feng,
  • Yue Lan,
  • Yanjie Zhao,
  • Wei Ke,
  • Yong Lin,
  • Kefeng Li

摘要

Autism spectrum disorder (ASD) affects tens of millions of families worldwide, yet parents confront abundant but unreliable online advice and limited access to timely, empathetic guidance. To address this critical gap, we developed Starmate (http://kefeng.mpu.edu.mo/starmate), a 1.5B-parameter, domain-tuned AI assistant for ASD caregivers, using a rigorous user-centered mixed-methods framework. Informed by in-depth interviews ( \(n=13\) ) and a Kano survey ( \(n=60\) ) that identified “Hands-on guidance” as a must-have caregiver requirement, we engineered a novel modular architecture that integrates sentiment analysis, expert-vetted knowledge-graph-augmented retrieval (LightRAG), and a domain-fine-tuned Qwen2.5-1.5B model. In a blinded, side-by-side comparison against leading commercial LLMs, Starmate demonstrated improved performance across key metrics within this evaluation framework (86.76 vs 78.43–83.84; \(p < 0.001\) ) and showed specific advantages in Empathy, Hands-on guidance, and Logical clarity (all \(p < 0.05\) ). Automated benchmarking corroborated these results, with top scores for Professional accuracy (86.18), Empathy (86.79), and Hands-on guidance (82.58). These findings demonstrate the technical feasibility of a lightweight, privacy-conscious, domain-specific LLM to generate accurate, empathetic, and actionable responses in benchmarked scenarios, laying the groundwork for future real-world usability and clinical testing.