Automated population reconstruction of the 19th century Tilburg region

Today marks the point at which my population reconstruction of the Tilburg region is finally online. Below a short summary of what happened and how it happened, and a link to the website: People of Yesterday. More updates will follow.

The enthusiasm of a teenager

In 2016 I completed my bachelor thesis in Computer Science at the University of Amsterdam. Driven by my interest for Dutch history, I analyzed the 19th century migration network around Tilburg and Rotterdam. To do this, I combined Dutch civil registration data and created individual timelines of events for all people who lived within the region. When I found two points in time that were registered in a different municipality, I assumed a migration event took place. Generally Dutch archivists were very enthusiastic about the way of research and new way of using historical data. Moreover, the results proved to be quite accurate.

Focus on quality

Flash forward more than 3 years. It is 2019 now, I have completed my master in Computational Science and have even recently started work in a PhD position at NIDI, studying Dutch divorce behavior. The one thing that kept me from publishing my database earlier on, has changed: I am now feeling more confident about the quality of my work. Last summer, I had some time off and decided to focus on getting that database out that had been sitting on my hard disk. I ate some chocolate, took a cup of tea and started work.

Since then I’ve greatly improved my name matching, and made sure to include references to records in the database as well… I made it easier for myself to approach the database, and rebuilt it. Twice. Once using data from the Regional Archives of Tilburg, and once using the data from the Brabant Historic Information Centre (BHIC). For now, I finally want to share with you the Regional Archives of Tilburg database – the BHIC doesn’t have images attached to all records.

But is it usable?

Before doing so, one word of caution: my computer has yet to develop a mind of it’s own to be smart enough to correct all mistakes in assumptions I made, which means that the database is not perfect. Sometimes it will miss a death date, and from what I saw the algorithm managed to combine two families into a family with 30 children in the BHIC database… But you know, it’s a start. In my opinion it’s a good start and hopefully something that is both inspiring and useful to others.

What do I still want to fix in the coming time? A better search function would be nice – you know, something which doesn’t require you to type in the exact same name as the database. Something with wildcards, or something smarter… I also think it might be nice to be able to log-in, save favorites and change your language settings to English! In addition to this, I really want the names to be inside the source citations. So much for short-to-medium term goals.

So, can I use it now?

Yes. You can. (In Dutch.) On People of Yesterday . Go on, try it! Search some names, see what comes up, and try to get used to see your family as a network. Maybe it can be of use in placing DNA matches, or to expand your tree. On the left side of the screen you will be able to click and see the actual records. Are you excited in a somewhat from a more scientific or demographic viewpoint? Send me a LinkedIn or Twitter message, and I hope to drink a nice cup of hot chocolate with you soon!

Leave a Reply

Your email address will not be published. Required fields are marked *