Join the Textual Studies Program at two events related to geospatial humanities here on the Seattle campus:
Maps as Text and Text as Maps
2:30-4:30 pm, Wednesday, November 2, 2022
University of Washington (Seattle campus), Communications Building CMU 202
Talk by Katie McDonough (Alan Turing Institute) followed by a tutorial with Ludovic Moncla (National Institute of Applied Sciences & LIRIS Laboratory).
Talk: Maps as Humanities Data
Katie McDonough, The Alan Turing Institute, London UK
We’ve had several years to consider what it means to have computational access to 1 million books. But what about maps? With so many images being scanned around the world, researchers can imagine using very large collections of digitized maps as primary sources. How can computational methods and the data they create transform the ways we search for and interpret information from the past? What does it mean to turn images into structured text data? In this talk, I explore how creating humanistic data from maps allows us to pursue creative spatial analysis.
Katie McDonough is a Senior Research Associate on the Living with Machines project at The Alan Turing Institute in London, UK, and, from January 2023, a Lecturer in Digital Humanities in the Department of History at the University of Lancaster. She completed her PhD in History at Stanford University and has held teaching and research positions in the US, Australia, and UK. Katie is a specialist of eighteenth-century France and works broadly on computational spatial approaches to early modern and modern history, including the GEODE project. Most recently, she has been PI of Machines Reading Maps, a transatlantic, interdisciplinary project developing methods to make text on maps useful data for humanities research.
Tutorial: Creating Geospatial Data from Historical Texts in French
Ludovic Moncla, National Institute of Applied Sciences (INSA) & LIRIS Laboratory (UMR 5205 CNRS), Lyon, France
In this tutorial, we demonstrate how to use a custom version of the Perdido geoparser python library. Using texts in French from Diderot and d’Alembert’s Encyclopédie as a case study for querying a corpus and wrangling geoparsed data, you will be able to compare Perdido’s Named Entity Recognition (NER) output to the results of other well-known NER libraries. In addition to the core elements below, we’ll discuss why text and spatial analysis can be difficult, but ultimately very rewarding with historical, non-English languages.
In this tutorial, we will demonstrate how to:
- Load data from TEI-XML files into a Python dataframe;
- Use a dataframe for simple data analysis;
- Test the Perdido Python library for geoparsing (geotagging + geocoding);
- Display geotagging results;
- And explore geocoding results on a map.
Ludovic Moncla is an Associate Professor at INSA Lyon since 2018 and is a member of the Data Mining & Machine Learning team at LIRIS Laboratory (UMR 5205 CNRS). He obtained a PhD in Computer Science in 2015 from University of Pau (France) and University of Zaragoza (Spain). His research interests include pluri-disciplinary aspects of Natural Language Processing, information retrieval, data mining, digital humanities and geographical information science. He is currently scientific manager on the interdisciplinary GEODE project (funded by LabEx CNRS ALSAN, 2020-2024) on the development of methods for diachronic study of geographical discourse within French encyclopedias.
Both speakers will be in Seattle for the ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems and the 6th ACM SIGSPATIAL International Workshop on Geospatial Humanities.