I am very excited to join the Ancient Identities project and, as a new team member, I would like to welcome everyone – hello!

I specialize in computational archaeology – a relatively young, but rapidly developing sub-discipline, which covers various applications of computational methods in archaeological and cultural heritage research.

To provide a few more details: the first attempts to use computers in archaeological analysis were made about 30 years ago and were largely related to the applications of Geographic Information Systems (GIS). While GIS remains an important part of computational archaeology, current approaches incorporate a much wider range of tools, such as computer simulation, 3D modelling, virtual reality and data mining.

So, how can these methods help us study the contemporary meanings of the past?

To illustrate, Chiara and I are currently exploring what users of social media have to say about the Roman period in Britain. Luckily, we have a huge amount of data available to us coming from Facebook pages and Twitter, for example. The amount of data available is actually so large and unstructured that it would be very difficult and extremely time-consuming to examine it manually… this is where data mining techniques come in handy.

We are retrieving relevant comments that were posted on public Facebook pages, while Mark Altaweel is helping us obtain similar data from Twitter. This task is significantly simplified by the fact that we are able to re-use some of the functions already developed for searching keywords, for example as part of the Rfacebook library.

Ancient Identities Facebook Page in the browser and extracted with Rfacebook.

(R is an open-source statistical package and programming language which allows users to add functionality – and, as a result, it has a large community of users sharing the codes they develop for free re-use. As a reminder, the code developed for Ancient Identities research can be found here).

The next and more challenging task ahead will be to find a way to verify, in an automated or semi-automated way, whether the text containing certain keywords (e.g. Roman Empire, Caesar, etc.) is actually relevant. Once this is done, we will be able to push the data to our database, which should be up on the server very soon, and to start organizing and analyzing the extracted data to understand how certain pasts are used in discussions related to present-day problems.