Argument Structure Mining in Scientific Articles: A Comparative Analysis

Scientific articles serve as examples to demonstrate the standard functional units of argumentation, extracting its argument structure would benefit the knowledge discovery, representation and retrieval. This study proposes a coding schema based on the argumentation ontologies, and uses the schema to annotate 40 articles in two research field, then combines with sequential pattern mining to find argument structure model. The results expose a similar linear construction with differences in the specific argument structures of each knowledge domain. Therefore, our models have strong applicability in these two fields.