Work

Mapping Heritage in South Asia: A Multi-Source Geospatial Framework for Spatially Explicit Documentation of Archaeological Sites using Machine Learning and Open-Source Data Platform (Arches)

Year: 2026

Type: conference-paper

Abstract: The Mapping Archaeological Heritage in South Asia (MAHSA) project has developed a multi‑source geospatial framework to improve documentation of archaeological heritage in South Asia. It integrates legacy datasets, georeferenced Survey of India historical maps, machine‑learning feature detection, and ground‑truthing field surveys into the Arches platform. This approach addresses long‑standing issues of spatial inaccuracy in older records and significantly enhances heritage site identification. Results show thousands of newly detected features and improved precision for legacy sites, visualised through bivariate mapping and the Arches platform. The methodology is transferable, supports SDG 11.4, and strengthens heritage management through reproducible, open‑science practices.

Source: Zenodo (CERN European Organization for Nuclear Research)

DOI: 10.5281/zenodo.19430281

Authors: Gerrits, Petrus J., Jabbar, Junaid Abdul, Campbell, Rosie, de Souza, Jonas Gregorio, Khan, Afifa S., Sarmah, Mou, Singh, Anisha, Suarez Moreno, Maria, Suganya, Kuili, Tomaney, Jack, Vafadari, Azadeh, Vidyarthi, Vaneshree, Roberts, Rebecca, Petrie, Cameron A.

Institutions: University of Cambridge, Vrije Universiteit Brussel, UCLouvain, Universitat Pompeu Fabra, Banaras Hindu University, European Commission, Joint Research Centre, Indian Institute of Technology Gandhinagar

Language: English

Topic: Archaeological Research and Protection

Subfield: Archeology

Field: Social Sciences

Domain: Social Sciences

Open Access status: green

OpenAlex ID: W7149902181

FWCI:

Cites:

Cited by:

Related to: 10