Human-Centered DNS Agent

: Kvalitetssäkring av registreringsdata

Abstract

Accurate registration data is essential for maintaining control over domain names. For domain name specialists, the process of reviewing and validating this data is often manual and time-consuming. Inaccuracies can lead to ownership disputes, brand damage, and reduced access to critical societal services. Despite the vital importance of data accuracy, deficiencies in registration data have been identified (ICANN 2024). At the same time, the General Data Protection Regulation Act (GDPR) has limited transparency in WHOIS data, complicating manual verification efforts. The Network and Information Systems Directive (NIS2) further imposes stricter requirements for verification and traceability (CENTR 2023). This study explores how a DNS agent, designed in accordance with the principles of Human-Centered AI (HCAI), can serve as a decision-support tool. The research integrates contemporary scientific theory in interaction design and artificial intelligence with practical design methodology. Through literature reviews, user interviews, expert consultations, and iterative prototyping, a DNS agent was developed and evaluated. The solution combines Conversational User Interfaces (CUI) and Visualization-Oriented Natural Language Interfaces (V-NLI) with Large Language Models (LLMs). The results indicate that a Human-Centered DNS agent has the potential to streamline the quality assurance of registration data while simultaneously enhancing user decision-making capabilities.

Date of Award	2025-Jun
Original language	Swedish
Supervisor	Kristoffer Åberg (Supervisor), Maria Freij (Supervisor) & Kari Rönkkö (Examiner)