2023年1月，数字科学促进与研究协会宣布其在线学术评审系统正式上线运行。这一系统的推出标志着协会成果评审工作从传统的邮件通信模式全面迈入数字化时代，大幅提升了评审效率和用户体验。
在线评审系统由数智研究院技术团队自主开发，覆盖了成果评审的全流程：课题申报提交、申报材料管理、评审专家自动匹配、匿名评审执行、评审意见反馈和评审结果公示。系统内置了智能化的专家匹配算法——结合提交成果的研究方向关键词、评审专家的研究背景和历史评审记录，自动推荐最匹配的评审专家组合，同时严格执行利益回避规则。
系统还引入了可视化的评审进度追踪功能。提交人可以实时了解自己的成果评审处于哪个阶段、预计完成时间，以及评审完成后的详细评审意见（匿名）。这一透明化设计体现了ADSPR"开放、严谨"的核心理念。
协会秘书长张瑞晨介绍，在线评审系统的开发历时八个月，经过了三轮内部测试和两轮会员测试。系统上线后，成果评审的平均周期从此前的60个工作日缩短至35个工作日左右，评审专家的匹配效率提升了近一倍。
值得一提的是，系统的自然语言处理模块由数智研究院评审专家孙婉清博士团队提供技术支持，能够对提交材料进行自动关键词提取和方向分类，辅助评审专家更快速地把握成果核心内容。
在线评审系统的上线也为2023年数字先锋奖的评选提供了更高效的技术支撑。第四届数字先锋奖的全部申报和评审工作将首次完全在线完成。
协会将持续优化系统功能，下一步计划增加多语言支持和移动端适配，进一步降低国际用户的使用门槛。
In January 2023, the Association for Digital Science Promotion and Research announced that its online academic review system had officially gone live. The launch of this system marks the Association's achievement review work moving fully from the traditional email communication model into the digital era, greatly improving review efficiency and user experience.
The online review system was independently developed by the technical team of the Digital Science Research Institute and covers the full process of achievement review: project application submission, application material management, automatic matching of review experts, anonymous review execution, review feedback, and publication of review results. The system includes an intelligent expert-matching algorithm that combines the research-direction keywords of submitted achievements, the research backgrounds of review experts, and historical review records to automatically recommend the most suitable expert groups, while strictly enforcing conflict-of-interest avoidance rules.
The system also introduces a visual review progress tracking function. Applicants can learn in real time which stage their achievement review is in, the estimated completion time, and the detailed anonymous review comments after the review is completed. This transparent design reflects ADSPR's core philosophy of "openness and rigor".
Secretary-General Zhang Ruichen explained that development of the online review system took eight months and went through three rounds of internal testing and two rounds of member testing. After the system went live, the average cycle for achievement review was shortened from the previous 60 working days to about 35 working days, and the matching efficiency for review experts nearly doubled.
It is worth noting that the system's natural language processing module was supported technically by the team of Dr. Sun Wanqing, a review expert at the Digital Science Research Institute. The module can automatically extract keywords and classify directions from submitted materials, helping review experts grasp the core content of achievements more quickly.
The launch of the online review system also provided more efficient technical support for the 2023 Digital Pioneer Awards selection. All applications and review work for the fourth Digital Pioneer Awards will be completed fully online for the first time.
The Association will continue to optimize system functions. The next plan is to add multilingual support and mobile adaptation, further lowering the usage threshold for international users.