GeoNewsMiner helps researchers with studying migration history

The new app GeoNewsMiner (GNM) visualizes georeferences in historical newspapers. The app, developed by researcher Dr Lorella Viola (Cultural History) and Prof. Jaap Verheul in collaboration with the Research Engineering Team of Utrecht University, adds a geodimension to the experiences of migrants.

Lorella Viola explains: "The app allows users to visualize geographical references in historical datasets. GNM displays places mentioned in Italian newspapers at the beginning of the last century. Because the dataset used is a collection of digitized immigrant newspapers, GNM will open up new ways to study migration history. Research in migration studies and diasporic cultures has shown how in diasporic contexts, migrants often mark their identity in connection with a place that is distant in time or space, typically the homeland. So we asked ourselves: what do diasporic references to places tell us about the way migrants interacted with the homeland and the host environment? Does visualising such references over time reveal complex mechanisms of interaction?”

Spatial humanities

“GNM positions itself within the recently emerged interdisciplinary field of spatial humanities” dr. Viola explains.  “Spatial humanities is a field which focusses on geographic and conceptual space, particularly from a historical perspective. Essentially, spatial humanities is based on Geographic Information Systems (GIS) which is used to georeference a dataset, map it, display it, and analyse it by cross-referencing different layers of information. Typically, the visualisation is done by using static maps, but the innovative part of GNM consists in the fact that, as visualisation tool, we chose to use an interactive, user-friendly app.”

Benefits of an app

Unlike many other methods in spatial humanities, the GeoNewsMiner allows the researcher to view all results in an interactive, intuitive and reproducible way. “Unique to this tool is the possibility to apply a wide range of filters like time range (1898-1920); newspaper’s title; least and most mentioned places; absolute or relative frequency; and aggregation on national, regional or city level. You can also overlay historical maps of selected years (1880, 1914, 1920, 1994), and download and share the data/results. All these features are indeed possible because we chose to use an app instead of static maps.”

Customizable preferences

“The great thing is that researchers can tailor their user experience based on individual research questions without being restricted too much by pre-determined methodological decisions that we as researchers made for our own project. Because that is a problem that many researchers in digital humanities face”, Dr Viola explains. “This digital method is also much faster to work with and makes it easier to share your data.”

We asked ourselves: what do diasporic references to places tell us about the way migrants interacted with the homeland and the host environment?

Step-by-step explanation

Lorella Viola hopes that the GeoNewsMiner can be used in several areas of interdisciplinary research: “One application of GNM is for example the potential to reconstruct the “geographical agenda” of historical newspapers by analysing potential“geographical biases” of the press. This is an issue urgent to fields such as media studies, cultural history and international relations. Moreover, we created a GitHub repository that describes in detail the step-by-step process of tagging the corpus, extracting the data, geocoding them, and creating the app. In this way, researchers will be able to reuse our methodology with their own dataset.”

YoDa

To store the data on which the app is based, Dr Viola used YoDa, a digital storage and collaboration system from Utrecht University: “Yoda was crucial for this project. YoDa satisfies the FAIR principle of proper data management: Findable, Accessible, Interoperable and Re-useable. If the corpus is published via YoDa, all metadata can be retrieved by various other catalogues. We in fact stored and published with YoDa both the datasets on which GNM is based: ChroniclItaly, the original collection of seven Italian immigrant newspapers, and ChroniclItaly 2.0, the version of the same collection tagged for entities, such as places. By using YoDa we advocate for a whole opening up of academia by sharing not just results, like in traditional scholarship, but also the resources and the tools to access them.”

GNM is the result of a fruitful collaboration with the Research Engineering Team of Utrecht University. Funding was provided through the Utrecht University Innovation Fund for Research in IT.