Universiteit Leiden

nl en

'Data science enables us to develop new tools'

PhD students Alex Brandsen and Wouter Verschoof-van der Vaart are both doing a project within the university’s Data Science research programme. The are introducing terms like ‘text mining’ and ‘advanced machine learning’ into archaeology. ‘These techniques will make archaeology more efficient and cheaper.’

Connected

Alex Brandsen is investigating how archaeological information in documents can become more accessible. ‘As we are using different terms for certain periods and areas, we are making it impossible to retrieve all the information. Also, the large diversity of resources is a problem. Just think of all 4000 archaeological research projects, like piston cores and excavations, in the Netherlands each year. They are being published in a great many systems, which you would really love to be interconnected. Then, you will be able to find a publication about a certain year, place or artefact, even if it is not mentioned in the abstract or as a key word. For quite often, in an excavation of the Middle Ages, material from the Roman Era is found, too. I hope that with my text mining research, I will be able to provide such a system.'

 

A bare earth model reveals the archaeological structures on the earth surface.

Computer recognizes structures

Brandsen’s colleague Wouter Verschoof-van der Vaart is occupied with totally different data-research. ‘I am looking at the automatic detection of archaeological remains with remote sensing. I am using images that have been taken from airplanes and satellites. With LiDAR – Light Detection And Ranging – you can make a model of the bare earth, without for example trees and bushes. With advanced machine learning, the computer recognizes the structures on the pictures. That is how, from a distance and on a much larger scale than right now, you would be able to recognize barrows, field systems and roads.

Profit

Associate professor Karsten Lambers is supervising PhD student Verschoof. He is convinced that archaeology will profit much from data science. ‘I believe that data science is offering us many chances. The new techniques will make our research area more efficient and cheaper. If you could photograph pottery on site and the computer system will determine the group to which the ornaments on the material belongs, that would save you a lot of work. You could then find characteristics and patterns that you might otherwise never find.’

Different resources

Lambers: ‘The Data Science Research Programme is offering us the chance to develop new tools that are specifically tailored for archaeological requirements. For a long time, archaeology was used to deal with scarce data. Data science makes it possible to efficiently and effectively analyse large amounts of data from different resources, that haven’t even been collected with archaeological applications in mind.’

Welcome to the start symposium

In 2017, 14 PhD students from all seven Leiden faculties started a data science project that concentrates on their specific research field. On Thursday 7 September, the Data Science Research Programme is organizing an opening symposium in the Academy Building. All employees of Leiden University are most welcome, specifically those who are interested in using data science for their own research in the future. Next year, another number of PhD positions will be granted.