Abstract: 【Purpose/significance】Analyze the knowledge discovery service driving model of smart libraries, and promote the develop?
ment of library data assets and the improvement of smart services.【Method/process】Based on data-driven, focusing on the five da?
ta-driven elements of user, content, analysis, service, and environment, analyze its data-driven process and dimensions, and establish
a power mechanism model through the interaction and correlation between driving dimensions.【Result/conclusion】From the perspec? tive of the operation of the power mechanism, data-driven is the flow of data, and the force relationships between different data-driven dimensions are all positive, through the data flow of data input, semantic association, visual analysis, feedback adjustment, data collab? oration with the basic database of the library, institutional repository, expert database, subject librarian database and other databases is realized, and pursue data intelligence in data collection methods, data mining techniques, data evaluation models, and hardware and software conditions.【Innovation/limitation】Data attributed dimensions are related to the realization of data collection, value transforma? tion power and the dimensions of subject intelligence determine the transformation and promotion of data to knowledge, the materiality dimension of data is the source to achieve the direct performance of the drive, and thus the feedback is related to the starting point and end point of the drive.