Research on argument mining has been a part of natural language processing (NLP) for more than ten years. Although it has potential applications in legal, political, and social contexts, this approach has largely been overlooked in organizational research. In their article, Dr. Gregor Wiedemann, Cornelia Fedtke, and Cristina Besio introduce aspect-based argument mining (ABAM) as a new method.
This article is the result of collaboration on the project “Few-Shot Learning for Automated Content Analysis in Communication Science (FLACA),” in which the HBI applied methodological insights gained through a collaboration with Prof. Besio’s department at Helmut Schmidt University/University of the Federal Armed Forces in Hamburg to the field of organizational research.
Abstract
Argument mining—the automatic identification, classification, and linking of argumentative text—has been studied in natural language processing (NLP) for more than a decade. Despite its claimed potential for applications in legal, political, and social contexts, it remained largely unexplored in organizational research. This article introduces aspect-based argument mining (ABAM) as a methodical innovation for studying how organizations justify decisions, construct legitimacy, and relate to their environments through communicative acts. By scaling up the analysis of argumentative structures beyond the limits of small-scale, qualitative studies, ABAM enables the recognition and systematic analysis of argumentation patterns in large text corpora that were hardly detectable with previous (computational) approaches. The potential is demonstrated by a longitudinal case study of Twitter debates on nuclear energy in Germany, revealing how shifting societal values—particularly the reframing of nuclear energy from a safety to a climate issue—produced growing misalignments between organizational talk of a political party organization and its social media environment.
Fedtke, Cornelia; Wiedemann, Gregor; Besio, Cristina (2026): Organizational Research Methods (29/3). https://doi.org/10.1177/10944281261453948