Our research group, ”Digital Lens” (dl.rc2s2.elte.hu) is part of the ELTE Research Center for Computational Social Science (rc2s2.elte.hu/en). We use computational tools, Natural Language Processing,
Our research group, ”Digital Lens” (dl.rc2s2.elte.hu) is part of the ELTE Research Center for Computational Social Science (rc2s2.elte.hu/en). We use computational tools, Natural Language Processing, machine learning and data visualization to analyze historical sources. We investigate written sources: early testimonies of Jewish Holocaust survivors recorded by the National Committee for Attending Deportees (DEGOB for short in Hungarian). We complement the computational analysis with qualitative methods. Our research questions address three aspects: gendered experiences, language usage and topography. We also deal with the overarching topic of ethics. The attached presentation focuses on topography.
Data visualization can be created driven by two purposes. The first is to explore data to find new research questions, get familiar with data, and find patterns. And the second is to explain the findings. Our charts are both exploratory and explanatory. The visualizations we create are not only illustrations but communicate knowledge. The charts in the slides are built around the same topic: the decentralization of Auschwitz. We show through different approaches how widespread the camp-system was: we use networks, flow charts and flow maps to present that Auschwitz was not the final destination of deportations as historiography on Hungarian Jews has so far suggested.
Our visualizations focus on those deported from only one specific ghetto: from Beregszász. The first chart, a snippet from our radial dendrogram created in d3 (a JavaScript library) shows the main camps and subcamps where Jews were deported to. Our second approach was “linear”. The second chart, called Sankey diagram shows the number of people with similar fates: those who had to forcibly do the same routes.
On the Tableau dashboard, we show two different layouts. This one is like the previous flow chart but shows geographical information also (if you pay close attention, you can see, that today’s borders and those from 1944 are present at the same time). When building this dashboard, we had some ethical considerations in mind: is it OK to show big numbers on our visualization? Should or shouldn’t we display names of survivors? Wouldn’t it be better to separate the lines, so that all fates can be present on the map simultaneously? We have decided to show numbers, names, and separate lines on the map. Through filtering you can see different fates of the survivors and get more familiar with them.
Click here to view our presentation