Horizon scanning helps organizations anticipate future developments and prepare for emerging risks. Traditional approaches, however, rely heavily on domain experts, making results difficult to scale, communicate, and reproduce. This research introduces an automated framework combining NLP and Large Language Models to replicate expert sensemaking. Unstructured texts are transformed into structured insights within a time-aware knowledge network, enabling automated detection of the knowledge origins, emerging topics and their evolution. LLM-based interpretation translates complex network results into understandable insights, making horizon scanning results accessible without requiring domain expertise.
Researcher
Publications
- Rubin, N., Haan, P., Blum, R. (2025) „Harnessing AI to Navigate User Generated Content: A Framework for Consistent Customer Experience Across Digital Touchpoints“, Presented at the International Marketing Trends Conference 2025 (IMTC)
- Haan, P., Berbig, M., Blum, R., Jörden, J., Schirrmeister, E., & Zimmermann, R. (2023). Mehrstufige strategische Frühaufklärung durch iterative automatisierte Themenerkennung und Fusion von Nachrichten-, Journal- und Patenttexten mittels Natural Language Processing (NLP). Presented at the 17th Symposium für Vorausschau und Technologieplanung (SVT). https://doi.org/10.24406/h-452213
- Haan, P., Klosa, O., & Nöth, F. (2023). VAG-Kundenmonitor – KI-gestütztes Feedback-Management für den Nürnberger Nahverkehr. DER NAHVERKEHR, 41(11), 34-38. https://eurailpress-archiv.de/SingleView.aspx?lng=de&show=5895418
Partner
