Abstract:

[Purpose/Significance] Research libraries constitute a vital component of China's science and technology (S&T) think tank ecosystem. Their intelligence services for government departments for S&T project performance evaluations can enhance the scientific rigor and accuracy of the outcomes while catalyzing their own diversified transformation and development. Existing research has demonstrated that research libraries can effectively support government departments in delivering intelligence services such as S&T project performance evaluation. However, current research predominantly focuses on holistic studies of intelligence services undertaken by research libraries, with a notable deficiency in targeted investigations specifically addressing their engagement in S&T project performance evaluation services. Anchored in the empowerment of multi-source data, this study focuses on exploring service models for the evaluation of S&T project performance in research libraries. It examines the potential of these services for government departments conducting multi-source data-empowered S&T project performance evaluation. Additionally, the study promotes the enrichment of vertical application scenarios in library and information science-related fields and supports the construction of digital and intelligent systems for S&T evaluation. [Method/Process] Through literature review, web-based investigation, and information correlation methods, this study systematically examines typical domestic and international practices of multi-source data-empowered S&T project performance evaluation, analyzes the alignment between such practices and research libraries' capabilities, and elucidates the significance of research libraries' engagement in S&T project performance evaluation. Based on the typology and characteristics of multi-source data, and leveraging the inherent advantages of research libraries, we clarify the evaluation rationale and principles, and construct a logical framework and innovative model for multi-source data-driven S&T project performance evaluation. [Results/Conclusions] Typical countries and regions both domestically and internationally emphasize the flexible application of multi-source data in project performance evaluation, establishing management platforms that aggregate full-lifecycle project data, and underscoring inter-departmental collaboration and data linkage among multiple government agencies. As neutral third-party institutions and critical data hubs, research libraries, when conducting S&T project performance evaluation under the new era context, are characterized by their emphasis on integrating fiscal fund effectiveness, S&T policy implementation efficacy, modernized governance systems, and multi-source S&T project data. They can construct innovative models for S&T project evaluation across four dimensions: data infrastructure development, indicator system construction, performance analysis and evaluation, and evaluation results application - thereby supporting the formation of a complete closed-loop chain of "evaluation-feedback-correction-improvement" for S&T project performance evaluation. Furthermore, this study proposes strategic recommendations for research libraries to broaden vertical application scenarios of S&T intelligence and deepen services for S&T project performance evaluation. These suggestions are listed as follows: strengthening policy intelligence service capabilities to enhance decision-support levels; deepening data resource integration and sharing to improve data support capacity; and strengthening the development of interdisciplinary talent teams to advance intelligence service quality. Future research directions will involve conducting specific case study analyses to provide support for refining the model of research library services for S&T project performance evaluation.

Key words: performance evaluation of science and technology projects, research libraries, innovation model, data base