Five Years of Argument Mining: a Data-driven Analysis

Abstract

Argument mining is the research area aiming at extracting natural language arguments and their relations from text, with the final goal of providing machine-processable structured data for computational models of argument. This research topic has started to attract the attention of a small community of researchers around 2014, and it is nowadays counted as one of the most promising research areas in Artificial Intelligence in terms of growing of the community, funded projects, and involvement of companies. In this paper, we present the argument mining tasks, and we discuss the obtained results in the area from a data-driven perspective. An open discussion highlights the main weaknesses suffered by the existing work in the literature, and proposes open challenges to be faced in the future.

Classifying argumentative stances of opposition using Tree Kernels

“…The main target of Argument Mining (AM) is to analyze arguments, including their components and the relations connecting these components [1] [2]. With an increasing number of works written and the interest of important private actors like IBM, this field has attracted a growing attention in the last few years [1], achieving important results and applications. These applications have been successfully implemented in a wide range of domains, since AM is physiologically multidisciplinary and facilitates cooperation among fields (e.g.…”

“…Cabrio and Villata [1] proposed a simpler two-step pipeline, which is the one that we will refer to in this work. In their pipeline, the first step is the identification of arguments, which involves not only the differentiation between argumentative and non-argumentative data but also the identification of the roles of argumentative components (claims, premises, etc.)…”

A Framework for Argument Retrieval

“…Computational argumentation is an emerging research area that has recently received increasing interest. It deals with representing and analysing arguments for controversial topics, which includes mining argument structures from large text corpora [8]. A widely accepted definition for an argument is that it consists of a claim or a standpoint, for instance "We should abandon fossil fuels", which is supported or attacked by at least one premise, for example "Burning fossil fuels is one cause for global warming" or "Poor people cannot afford alternative fuels" [21].…”

“…By now, existing (Web) search engines like Google only provide the most relevant documents to the user, but cannot structure their results in terms of claims and premises. There is a relatively large body of work on how arguments can be mined from text (see [8] for a recent survey). In this paper, we build upon established research on argument search engines and focus on effectively retrieving premises for a query claim from a large corpus of already mined arguments.…”

The evolution of argumentation mining: From models to social media and emerging tools

“…We provide the necessary background for argumentation schemes and we evaluate them based on their suitability on social media text, but we do not devote the entire paper to this topic. Both [37] and [38], surveyed a big spectrum of the AM field, describing models, corpora and methods, but they overlook the special nature of social me-dia and the special features they present. On contrary, we focus on text derived from social media and our entire approach is based on the characteristics of the social media; chaotic nature, noisy text, vague claims, complicated network relations, implicit premises, etc.…”