A Systematic Review of Applied Nonlinearity in Modeling Information Service Dynamics for Academic Library and Research Data System Networks

Abstract

Scholarly libraries and research data networks are becoming systems of networked information-related services that experience variable workloads, dynamic user demand, distributed resources, and the multifaceted interactions among repositories, servers and communication nodes. These characteristics provoke the application of nonlinear modeling to understand the stability of the system, its complexity, connectivity and service performance. This systematic review is going to examine how the dynamics of information-service in an academic library and a research data context can be modeled based on nonlinear concepts and network-based analysis. The analysis of relevant studies is done via a systematic review process, with the main focus on six indicators core metrics (Lyapunov Exponent, Entropy, Network Density, Average Path Length, Response Time, and Resource Utilization). The Lyapunov-based analysis is useful in assessing the stability and sensitivity of changing system states in terms of uncertainty and complexity of service demand and information flow which is the entropy. The density of networks and the average path length can help to assess structural connectivity and information spreading, but response time and resource usage are indicators of operational efficiency at different workloads. Existing studies are inclined to treat these dimensions in isolation, which creates a big gap in the combined study of nonlinear-network-performance. The review concludes that a combination of these measures might help to obtain a more impartial vision of the dynamic behavior of information-service and may help to design effective, strong and flexible networks of academic library and research data systems.