A Chatbot for Asylum-Seeking Migrants in Europe

TL;DR

ACME tackles the complexity of international protection law by delivering an explainable, auditable guide for asylum seekers. It combines a neural NL interface with a symbolic Abstract Argumentation Framework and a formal knowledge base to reason about protection options, using embeddings and LLM fallbacks for natural language understanding while preserving privacy. In validation with domain experts, ACME perfectly matched expert judgments across 10 cases, outperforming a GPT-4o baseline that succeeded only 3/10 and exhibited inconsistencies. The work demonstrates potential to reduce loads on migration authorities and humanitarian organizations while offering a transparent, privacy-conscious tool to support migrants before formal asylum applications.

Abstract

We present ACME: A Chatbot for asylum-seeking Migrants in Europe. ACME relies on computational argumentation and aims to help migrants identify the highest level of protection they can apply for. This would contribute to a more sustainable migration by reducing the load on territorial commissions, Courts, and humanitarian organizations supporting asylum applicants. We describe the background context, system architecture, underlying technologies, and a case study used to validate the tool with domain experts.

Figures (2)

• Figure 1: Overview of the interface.
• Figure 2: An excerpt of our Knowledge Base