Automatic Debate Evaluation with Argumentation Semantics and Natural Language Argument Graph Networks
This work addresses the lack of annotated data for complex NLP tasks like debate evaluation, though it appears incremental as it builds on existing argumentation and neural network techniques.
The paper tackles the problem of automatic debate evaluation by proposing a hybrid method that combines argumentation theory concepts with Transformer-based architectures and neural graph networks, achieving promising results that establish a foundation for analyzing natural language arguments.
The lack of annotated data on professional argumentation and complete argumentative debates has led to the oversimplification and the inability of approaching more complex natural language processing tasks. Such is the case of the automatic debate evaluation. In this paper, we propose an original hybrid method to automatically evaluate argumentative debates. For that purpose, we combine concepts from argumentation theory such as argumentation frameworks and semantics, with Transformer-based architectures and neural graph networks. Furthermore, we obtain promising results that lay the basis on an unexplored new instance of the automatic analysis of natural language arguments.