Automated Label Placement on Maps via Large Language Models DOI: XXXXXXX.XXXXXXX Conference: 1st Workshop on AI for Data Editing at KDD’25; August 3, 2025; Toronto, ON, Canada DOI: XXXXXXX.XXXXXXX ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Computing methodologies Machine learning

Harry Shomer, Jiejun Xu · 2025

Read Automated Label Placement on Maps via Large Language Models DOI: XXXXXXX.XXXXXXX Conference: 1st Workshop on AI for Data Editing at KDD’25; August 3, 2025; Toronto, ON, Canada DOI: XXXXXXX.XXXXXXX ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Computing methodologies Machine learning by Harry Shomer, Jiejun Xu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-22952.txt.

Description

Harry Shomer and Jiejun Xu automated map label placement in two thousand twenty-five. Labels make maps readable, but placing them is slow handwork. Rules for good placement depend on feature type and context. Old automated systems struggle to apply those rules. The pair recast placement as a data editing task. Language models supply the needed context awareness.

They built MAPLE, a benchmark of real-world maps. It covers varied landmarks with human placement notes. Their method fetches guidelines for each landmark type. Retrieval supplies those rules inside the model prompt. A tuned model then places labels to match them. The benchmark lets others score placement quality fairly. Map platforms could scale labeling to huge atlases.

Catalog record

CREATORHarry Shomer, Jiejun Xu
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE44 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-22952

Browse the full library · About the Tower of Babel