Digitized Knowledge-Based Literature Reviewing: A Tutorial on Coding Causal and Process Models as Graphs

Abstract

Current literature reviewing approaches rely on using keywords to search for relevant articles, reading them, and manually synthesizing their knowledge. This approach is inefficient and limits the pace of scientific progress. To address such issues, we have developed a graph-based approach to digitizing knowledge. The method focuses on coding the core knowledge in publications (i.e., causal or process models) as directed graphs following a well-defined regimen and combining the directed graphs into a labeled property graph database. We provide guidance on coding publications as graphs using a graph query language. An application has been developed to facilitate scholars to code publications as graphs to load into a graph database. We discuss the contributions of the method for literature reviewing.

Keywords: Knowledge digitization, Literature review, Causal model, Graph database, Graph query language