Nestedness Temperature in the Agent-Artifact Space: Emergence of Hierarchical Order in the 2000–2014 Photonics Techno-Economic Complex System

In this work we represent a techno-economic complex system based on the agent-artifact space theoretical model. The objective is to structure a methodology to statistically investigate the presence of hierarchical order, as an emerging property of this system. To analyse the agent-artifact space, two statistical methodologies are initially employed. The first is a community detection method, employed with the objective to detect groups of agents that are likely to intensively exchange information within the considered complex system. The second is a natural language processing method, the LDA topic model, employed with the objective of identifying types of artifacts as technological subdomains through textual information that describes the activities of agents. After this initial part, we address the investigation of the structure of the agent-artifact space by estimating the involvement of each community in the detected topics. This is effectuated by means of a statistic that considers the information flow percentage of agents, the fractional count of activities, and the probability of agents’ activities to belong to topics. We then estimate the hierarchical order of the topics’ distribution in communities, by computing its nestedness temperature, which is adopted by studies on ecological systems. This statistic’s significance is finally evaluated with z-scores based on homogeneous systems. The case study is a system consisted of economic agents (e.g. firms, universities, governmental institutions) patenting in the technological domain of photonics. The analysis is effectuated over five time spans in the period 2000–2014. The observed values of nestedness temperature are proved statistically significant, which suggests that hierarchical order is an emerging property of the agent-artifact space.