This technical report discusses the potential use of Large Language Models for location detection with a focus on the JRC Terrorism and Extremism Database. The report highlights the current inaccuracies in the database's location detection algorithm, which uses adhoc created embeddings, struggles with contextual issues, and has difficulty with translated location names. These issues can lead to misclassified events, negatively impacting the quality of the tool. The report suggests exploring more accurate and sophisticated approaches, taking advantage of recent advancements in Artificial Intelligence, to increase the accuracy of the automated classification thus helping reducing human intervention.
BOSSO Francesco;
VALISA Jessica;
RUBERTO Stefano;
VALSAMOS Georgios;
SCHUMACHER Ralf;
LARCHER Martin;
2023-12-19
Publications Office of the European Union
JRC134961
978-92-68-09641-3 (online),
1831-9424 (online),
EUR 31745 EN,
OP KJ-NA-31-745-EN-N (online),
https://publications.jrc.ec.europa.eu/repository/handle/JRC134961,
10.2760/575493 (online),
| Name | Country | City | Type |
|---|
This document is only visible at the Commission level.
You are not authorized to publish or distribute it outside the European Commission.
This is a public document. You can share this publication.
Datasets
| ID | Title | Public URL |
|---|
Dataset collections
| ID | Acronym | Title | Public URL |
|---|
Scripts / source codes
| Description | Public URL |
|---|
Additional supporting files
| File name | Description | File type |
|---|