mjstoroz / archaeology-machine-learning

machine learning techniques for archaeology.

👋 welcome to the archaeology machine learning repository

📖 introduction to the project

Machine learning (ML) techniques present new ways of approaching archaeological research questions and interest in applying these methods continues to grow. This repository documents the application of ML techniques to archaeological data, aiming to assist those working in the field by:

• providing an overview of the ways ML is being applied in archaeology
• sparking new ideas whilst reducing duplication of work
• encouraging the sharing of code, data, and other resources
• making resources more FAIR (Findable, Accessible, Interoperable, and Reuseable)

By doing this, we hope to support practitioners to learn about, critically apply, or contribute to conversations about, machine learning techniques for archaeology.

✅ how to contribute

This project is open for contributions!

Check out our 🗺️ roadmap to get an overview of the current milestones we're working towards and find out how to participate.

🗞️ releases

Archived releases of this repository with a citeable DOI will be made at regular intervals.

🙏 acknowledgements

This project was inspired by the satellite-image-deep-learning and AncientMetagenomeDir projects, and was kicked off as part of Open Seeds cohort 8.

📁 repository contents

The repository is structured into sections by application area (e.g. remote sensing), and then by machine learning technique, with linked examples illustrating the uses of each technique. Use the contents list below ⬇️ to browse the application areas included so far and jump to specific sections, or scroll down to view everything.

🛰️ remote sensing

1. classification
2. segmentation

📚️ textual analysis

1. language models
2. named entity recognition

---

🛰️ remote sensing

classification

title	authors	year	data type	technique	paper	code	data
Using CarcassonNet to automatically detect and trace hollow roads in LiDAR data from the Netherlands	Verschoof-van der Vaart and Landauer	2021	lidar	classification	paper	code:tbc	data:tbc

segmentation

title	authors	year	data type	technique	paper	code	data
Investigating ancient agricultural field systems in Sweden from airborne LIDAR data by using convolutional neural network	Küçükdemirci et al	2022	lidar	segmentation	paper	code:tbc	data:tbc

📚️ textual analysis

language models

title	authors	year	type	technique	language	paper	model
ArchaeoBERT	Alex Brandsen	2023	BERT	Masked Language Modelling	English	paper	model
ArchaeoBERT-NER	Alex Brandsen	2023	BERT	Named Entity Recognition	English	paper	model
ArcheoBERTje	Alex Brandsen	2023	BERT	Masked Language Modelling	Dutch	paper	model
ArcheoBERTje-NER	Alex Brandsen	2023	BERT	Named Entity Recognition	Dutch	paper	model
bert-base-german-cased-archaeo	Alex Brandsen	2023	BERT	Masked Language Modelling	German	paper	model
bert-base-german-cased-archaeo-NER	Alex Brandsen	2023	BERT	Named Entity Recognition	German	paper	model

named entity recognition

title	authors	year	type	format	language	paper	data
Creating a Dataset for Named Entity Recognition in the Archaeology Domain	Brandsen et al.	2020	data	CoNNL	Dutch	paper	data

🌏 site distribution modelling

regression

title	authors	year	data type	technique	paper	code	data
A supervised machine-learning approach towards geochemical predictive modelling in archaeology	Oonk and Spijker	2015	soil geochemistry	k-nearest neighbours, suppert vector machines, and artificial neural networks	paper	code:tbc	data:tbc
Mapping Tasmania's cultural landscapes: Using habitat suitability modelling of archaeological sites as a landscape history tool	Jones et al.	2019	climate and topography	Random Forest	paper	code:tbc	data:tbc
Advancing predictive modeling in archaeology: An evaluation of regression and machine learning methods on the Grand Staircase-Escalante National Monument	Yaworsky et al.	2020	climate, topography, and environmental productivity	Random Forest	paper	code	data
An Explorative Application of Random Forest Algorithm for Archaeological Predictive Modeling. A Swiss Case Study	Castiello and Tonini	2021	topography and soil characteristics	Random Forest	paper	code:tbc	data:tbc

⚛️ stable isotope analysis

regression

title	authors	year	data type	technique	paper	code	data
A bioavailable strontium isoscape for Western Europe: A machine learning approach	Bataille et al.	2018	strontium	Random Forest	paper	code	data:tbc
Advances in global bioavailable strontium isoscapes	Bataille et al.	2020	strontium	Random Forest	paper	code	data
A bio-available strontium isoscape for eastern Beringia: a tool for tracking landscape use of Pleistocene megafauna	Funck et al.	2020	strontium	Random Forest	paper	code:tbc	data