ReCAP-II – Information Retrieval and Case-Based Reasoning for Robust Deliberation and Synthesis of Arguments in the Political Discourse

The ReCAP–II proposal continues the research performed in the ReCAP project, currently funded within the RATIO DFG Priority Programme. ReCAP–II aims at significantly contributing to the foundations for building argumentation machines according to the overall goal of the Priority Programme. The project focus remains on argumentation to support researchers, journalistic writers, as well as human decision makers to obtain a comprehensive overview of current arguments and opinions related to a certain topic. Such argumentation machines automatically explore and process available information sources on the Web, particularly argumentative texts and factual content relevant for the specific topic under discussion. Unlike existing search engines, which primarily operate on the textual level, such argumentation machines reason on the knowledge level formed by arguments and argument structures. For a given particular context, such reasoning will actively support the deliberation of arguments and counter-arguments for the issue under consideration.  In addition, it will support the synthesis of new arguments, based on analogical transfer from similar related contexts and topics. For this purpose, the project aims at novel contributions to and confluence of methods from information retrieval (IR) and knowledge representation and reasoning, in particular case-based reasoning (CBR), for building argumentation machines. The aim is to develop methods that are able to capture arguments in a robust and scalable manner, in particular representing, contextualising, and aggregating arguments and making them available to a user. 

In particular, RECAP-II focusses on two major goals. The first goal is to further improve the methods already developed in RECAP with respect to specific challenges. In particular, we will address interactive, explainable CBR for argument synthesis as well as argument validation and evaluation. The second goal is to achieve, the integration of the developed methods in order to approach the overall goal of RECAP and the RATIO Priority Program, namely to build an argumentation machine. For this purpose, we will additionally investigate state-of-the-art methods from argument mining for extracting argument graphs from German texts as well as methods for user interaction, visualization, and context representation.  As a major result, we aim at developing a technical architecture of argumentation machines with clearly specified services and interfaces. Based on this, several use cases will be implemented as demonstrator application, which will then allow to perform an end-to-end evaluation using real users from journalism and political research.