Comment Analyzer

Abstract

The lack of visually guided data exploration tools limits the scope of research questions communication scientists are able to study. The Comment Analyzer addresses this gap by supporting the analysis of commenting behavior in news articles. Its comment-thread corpora contain comments tagged with deliberative quality indicators and political stance.

The tool provides a visual querying system for exploring and analyzing these corpora. It helps social scientists investigate the distributions and relations between comment attributes, the homogeneity of thread sets, frequent thread structures, and changes in comment qualities across individual and multiple threads. The system was developed in close collaboration with communication scientists using a user-centered approach. Reviews with communication scientists and five independent experts confirmed its utility for corroborating findings, answering new research questions, and supporting the analysis of other datasets in the communication sciences.

Publication

D. Kiesel, P. Riehmann, I. Engelmann, H. Ramezani and B. Froehlich. Comment Analyzer: A Tool for Analyzing Comment Sets and Thread Structures of News Articles IEEE Transactions on Visualization and Computer Graphics, vol. 27, no. 2, pp. 1139-1148, February 2021. DOI, Preprint, Video