Application of artificial intelligence technologies for optimizing process management based on metadata processing statistics in an electronic document management system in an educational organization
Abstract. Document automation is a key element of digital transformation, especially in educational institutions, where the volume of documents includes academic and administrative materials. Companies typically choose between managing the document lifecycle and automating business processes, where documents play a supporting role. In the educational sector, the second approach is more relevant, as it helps reduce approval times and improve data management efficiency. ECM systems have already become an integral part of university infrastructure, managing curricula, student applications, research publications, and financial reports. However, challenges such as semantic ambiguity, lack of standardization, and document duplication require new solutions. The integration of Al into ECM systems opens up opportunities for automatic classification, analysis of unstructured data, and optimization of approval processes. Nevertheless, the adoption of Al in education is progressing slowly because of technical and organizational barriers. The article investigates document automation in universities using the example of the student withdrawal process. Key approval stages are analyzed, bottlenecks are identified, and solutions are proposed. Special attention is given to the use of machine learning for metadata analysis, which enables the prediction of document processing routes and optimization of time expenditures. The aim of the article is to propose a document management efficiency model that considers both time and quality parameters. The model identifies "risk points," suggests improvement measures, and ensures effective document management. The research results can be used to enhance ECM systems in universities, particularly during periods of high workload.
Keywords: document automation, ECM systems, document lifecycle, business processes, artificial intelligence (Al), structured and unstructured data, optimization of approval processes, metadata and machine learning, digitalization of higher education, management of student body movement.
Highlights:
- key issues in document management in educational organizations have been identified, such as semantic ambiguity, lack of standardization, document duplication, and delays in approval, which reduce process efficiency;
- a document management efficiency model has been developed, based on the analysis of time expenditures, completeness of execution, and document quality, enablingthe identification of "bottlenecks" and process optimization;
- the use of machine learning has been proposed to automate metadata analysis, predict document processing routes, and reduce the workload on employees;
- the integration of Al into ECM systems improves the processing of unstructured data, such as research papers and educational materials, enhancing the value of analytics for educational processes;
- recommendations for process improvement have been developed, including workload distribution among employees, error elimination, and enhancing staff training for working with ECM systems.
Dmitry N. Frantasov, Elena V. Voronina - Samara State University of Economics, Samara, Russia