Argument exploration made simple with natural language interface
ABDA-NL: A Natural-Language Scenario Explorer for Argument-Based Reasoning
Artificial IntelligenceHuman-Computer Interaction
Summary
When people discuss complex topics, they often use arguments for and against ideas. The authors introduce a system that helps users see which arguments stand, which don’t, and which are unclear, all in regular English. Users can ask questions, try out different assumptions, and even add new facts or rules, with help from a language AI that turns plain English into formal logic. The core argument system then checks all changes to make sure they make sense before applying them.
What this means in practice
- •For legal teams: Help legal professionals explore different case arguments interactively using natural language input and formal logic validation.$Commercial implications: Enables development of legal argument analysis tools that let lawyers experiment with case scenarios in plain English, ensuring sound reasoning.
- •For policy advisors: Assist policy advisors in exploring implications of different assumptions and rules within policy discussions by providing an interactive argument scenario explorer.
Authors
Shawn Bowers, Martin Caminada, Haoyang Liu, Bertram Ludäscher
Abstract
ABDA-NL adds a natural-language interface to ABDA, a system for argument-based discussion using ASPIC- knowledge bases under grounded semantics. Users see which conclusions are accepted, rejected, or undecided, open an interactive rendering of the grounded discussion game to learn why, explore what-if alternatives by suspending assumptions and rules or changing preferences, ask questions that are answered from a scenario's reference documents, and author new facts, assumptions, and rules in plain English. A large language model provides the bridge between language and formalism: it answers questions from the documents and the current state of the scenario, and it translates plain-English edits into candidate formal statements. The deterministic ABDA engine remains the sole source of arguments, attacks, and acceptance labels, and every proposal of the model is validated and confirmed by the user before it takes effect.