Methodology to promote inter- and transdisciplinary knowledge integration by identifying the functional concept using network analysis

This is the authors' abstract. We don't add key points for this paper.

This study proposes a methodology for visualising the importance and relationships of ideas expressed in dialogue through network analysis based on a structured workshop process. Node centrality is evaluated from a directed graph that represents multiple sticky notes on an online whiteboard, where nodes represent these sticky notes and edges connect them. The purpose of the study is to identify functional concepts and clarify the flow of ideas during dialogue, thereby providing insight into how participants influence each other's thinking. Analysis of the relationships between the Answer nodes and directly connected nodes shows that centrality scores of nodes influencing the Answer nodes are generally high. Additionally, nodes in the dialogue exhibit increased PageRank, indicating proximity to highly linked nodes, alongside rising betweenness centrality, suggesting a correlation between these centrality scores. However, natural language processing analysis of transcripts reveals that the words on sticky notes of high-centrality nodes hold extremely low importance, contrasting with centrality analysis results. These findings suggest that conventional frequency-based text analysis may overlook influential ideas in dialogue, highlighting the added value of network-based evaluation. This study offers a new perspective for identifying functional concepts by visualising thought flow in dialogues, particularly as a practical learning tool in transdisciplinary dialogue for effective consensus-building among diverse stakeholders.