A Corpus of Argument Networks: Using Graph Properties to Analyse Divisive Issues

A Corpus of Argument Networks: Using Graph Properties to Analyse Divisive Issues Barbara Konat1, John Lawrence1, Joonsuk Park2, Katarzyna Budzynska3,1, Chris Reed1 1Centre for Argument Technology, University of Dundee, Scotland 2Department of Computer Science, Cornell University, New York, USA 3Institute of Philosophy and Sociology, Polish Academy of Sciences, Poland Abstract Governments are increasingly utilising online platforms in order to engage with, and ascertain the opinions of, their citizens. Whilst policy makers could potentially benefit from such enormous feedback from society, they first face the challenge of making sense out of the large volumes of data produced. This creates a demand for tools and technologies which will enable governments to quickly and thoroughly digest the points being made and to respond accordingly. By determining the argumentative and dialogical structures contained within a debate, we are able to determine the issues which are divisive and those which attract agreement. This paper proposes a method of graph-based analytics which uses properties of graphs representing networks of arguments pro- & con- in order to automatically analyse issues which divide citizens about new regulations. By future application of the most recent advances in argument mining, the results reported here will have a chance to scale up to enable sense-making of the vast amount of feedback received from citizens on directions that policy should take. Keywords: Argument Analysis, Argument Structure, Automatic Text Interpretation, Divisiveness Measures, e-Participation, Graph- based Analytics 1. Introduction The aim of this paper is to show how the analysis of argu- mentative and dialogical structures, and in particular pro- and con-arguments, allows for the principled identification of divisive issues in an online corpus of debates. More pre- cisely, we show that structuring data as networks of inter- acting arguments, claims, reasons and conflicts means that the data is represented in the form of a graph. This repre- sentation, in turn, offers the possibility of using graph prop- erties to automatically analyse which issues attracted most attention and which most divided disputants. We present an annotated corpus which builds on previ- ous work in argument mining (Park and Cardie, 2014) by adding further mark-up capturing dialogical interactions between participants in an eRulemaking system for online deliberative democracy. We demonstrate how the rich net- work structure available in this style of annotation supports multiple interpretations of divisiveness which in turn pro- vide powerful insights into the nature of the debate – in- sights which can be valuable in creating summaries and overviews of complex, multi-faceted debates for decision- makers who need sense-making tools to develop a coherent understanding of large volumes of data from public contri- butions – in this case on contentious policy issues within the remit of the Department of Transportation in the US. The paper is structured as follows. Section 2 discusses some related work in argument analysis and controversy detection. Section 3 presents language resources used in this study. Section 4 demonstrates the type of informa- tion which can be extracted by applying our model – we show how the structure …

Content truncated to 500 words