| 王明亮 陈实.数智工具在科技成果转化服务中的典型应用与辅助决策效能优化研究[J].中国科技资源导刊,2026,(3):52~60 |
| 数智工具在科技成果转化服务中的典型应用与辅助决策效能优化研究 |
| Typical Applications and Auxiliary Decision-Making Effectiveness Optimization of Digital-Intelligent Tools in Technology Transfer Services |
| 投稿时间：2026-04-14 |
| DOI： |
| 中文关键词: 数智工具;科技成果转化服务;人工智能;人机协同;辅助决策效能 |
| 英文关键词: digital-intelligent tools, technology transfer services, artificial intelligence, human-machine collaboration, auxiliary decision-making effectiveness |
| 基金项目:新疆生产建设兵团科技计划项目“兵团科技成果转化与技术转移平台优化及示范应用”(2025YD043)。 |
| 作者 | 单位 | | 王明亮 陈实 | (新疆生产建设兵团科技发展促进中心，新疆乌鲁木齐 830002) |
|
| 摘要点击次数: 22 |
| 全文下载次数: 15 |
| 中文摘要: |
| 聚焦科技成果转化服务的前端和中端辅助决策环节，围绕成果评价、报告生成、供需匹配建议、转化路径导航以及企业需求挖掘等场景，归纳数智工具的主要应用方式及其作用边界。通过文献梳理、问卷调查、案例回顾及对比实验发现，现阶段数智工具在底层数据、算法可解释性和输出结果质量等方面仍存在不足。基于此，提出人机协同增强型成果转化模型(H-CETM)，构建“智能输入—算法处理—人工校验—动态反馈”的运行机制，并选取120项科技成果开展配对对比验证。结果显示，相较于纯AI工具，H-CETM在内容准确性、分析深度、推荐合理性以及综合得分上分别提升25.4%、42.6%、37.1%和30.1%。研究表明，人机协同有助于提升数智工具在成果评价、报告生成和转化路径建议等服务环节中的辅助决策效能，但其作用边界主要在过程支持和前期判断，不能替代专家团队对最终产业化决策的深度论证。 |
| 英文摘要: |
| This paper focuses on front-end and mid-end auxiliary decision-making services in technology transfer, including achievement evaluation, report generation, supply-demand matching suggestions, transformation path navigation, and enterprise demand mining. It summarizes the major application modes, functions, and boundaries of digital-intelligent tools. Based on literature review, questionnaire survey, case review, and comparative experiments, the study finds that current digital-intelligent tools still have limitations in underlying data, algorithm interpretability, and output quality. On this basis, a human-machine collaborative enhanced technology transfer model (H-CETM) is proposed. The model establishes an operating mechanism of “intelligent input—algorithm processing—human verification—dynamic feedback” and is validated through paired comparison using 120 scientific and technological achievements. Compared with pure AI tools, H-CETM improves content accuracy, analysis depth, recommendation rationality, and overall score by 25.4%, 42.6%, 37.1%, and 30.1%, respectively. The results indicate that human-machine collaboration can improve the auxiliary decision-making effectiveness of digital-intelligent tools in achievement evaluation, report generation, and transformation path recommendation. However, its boundary lies mainly in process support and preliminary judgment, and it cannot replace expert teams in final industrialization decisions. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
