Network Analysis and Text Mining Approach for Moorland Systems Management and Self-Governance

Network theory is a widely used approach for characterizing the interaction between the components of complex systems. However, dealing with social, cultural and socio-ecological complex systems is challenging in their own particular ways. Interactions in such systems cannot be characterized entirely as socio-ecological because they are also culturally charged and semantically loaded from the perspective of their components. In order to achieve a better understanding, the component's view and symbolic interpretation cannot be ignored. Due to these singular properties, we believe that for the analysis, management and decision-making regarding these kind of systems, complementary and alternative analysis rooted in text mining is required. In such a way, all stakeholder's discourses are the base to capture additional information about the context-related properties of the interactions under study. This paper explores the integration of different computational techniques consisting of text mining and network analysis procedures as tools for the characterization of actors in the context of environmental resource management and governance. We develop a comprehensive methodology for describing the topology of the perceived interactions among stakeholders, characterizing them with the content of their discourses about moorland conservation. On one hand, the methodological approach permitted us to obtain: (i) the networks or “ego-centered networks” based on the stakeholders or actors interviews, (ii) a complete network by merging ego-centered networks, (iii) synergistic/antagonistic perceptions towards actors were represented in weighted networks, and (iv) the identification of leaders among the interviewees. On the other, text mining also permitted the analysis of interviews performed with actors, providing social and semantic context about the moorland conservation discourses. We finally integrated both approaches in a single framework. We believe that in using such methods, recommendations for self-governance may be possible. The results show the social structure of the moorland's leaders, their cultural context and their perceptions about the importance of the moorland ecosystem. Based on these findings, decision and policy-making for institutional and stakeholder actors could be improved. Moreover, we would like to suggest that in having the right description of the system, a guided self-organization approach could be applied