21 October 2020, 09:00 – 11:00 SGT
What technical tools unlock the content of maps and enhance interoperability?
Historical maps are an important primary source in aiding our understanding of geographical, urban, and colonial histories. How have researchers used data science, Geographical Information Science (GIS), International Image Interoperability Framework (IIIF) and digital technology (Optical Character Recognition and computerised visual analysis) to “unlock” historical maps, that allow us to understand historical developments in boundary changes, urban development, and land use?
Chair: Associate Professor Andrea Nanetti
School of Art, Design and Media
Nanyang Technological University
Associate Professor Yao-Yi Chiang
Associate Professor (Research) of Spatial Sciences, Spatial Science Institute; Associate Director of Integrated Media Systems Center, Viterbi School of Engineering
University of Southern California
Associate Professor Stefan Leyk
Associate Professor of Geography (GIScience)
University of Colorado, Boulder
Linked maps: exploiting context in cartographic evolutionary documents to extract and build linked spatial-temporal data sets
This talk will present our recent progress in extracting historical map contents and linking them to other knowledge bases. Specifically, this talk will cover automated approaches for 1) feature extraction from historical map images and 2) linked spatiotemporal data generation from historical map contents, including geographic features and text labels.
Mr Jack Reed
Geospatial Web Engineer
Digital Library Systems and Services
Stanford University Library
Building, aggregating, and using historical map collections with open technologies
This talk will highlight the technology and collaboration advances that have enabled new types of scholarship and use of geospatial data. Discussed topics will include: aggregating data sources for federated discovery using OpenGeoMetadata and GeoBlacklight, IIIF as a platform for utilising map images, and current and future work to extract data from maps.
This talk will give a general overview of the landscape of aggregating geospatial data and utilising technologies to extract more information.
Dr Katie McDonough
Senior Research Associate
The Alan Turing Institute
Maps as data: a humanistic approach to computer vision for large map collections
If humanities scholars have been slow to take up large-scale analysis of digitised maps, the scientific and commercial community has not. Scientific treatment of historical maps as uncritical proxies for a ground truth, or accurate representations of the built and natural environments threatens to reproduce inequalities and violences embedded in these sources. It is time for historians to develop humanistic methods for working with maps as data. In this talk, I demonstrate some work completed in the Living with Machines digital history project to use new computer vision methods to ask questions about industrialisation from large corpora of nineteenth-century British maps. In conclusion, I explain some of the obstacles to working with maps at scale and what steps we might take in the spatial humanities and GLAM communities to democratise digital spatial history that uses maps as primary sources.
About the speakers
Assoc Prof Stefan Leyk specialises in Geographic Information System (GIS) where his research interests lie in cartographic pattern recognition from historical maps, spatial dynamic modelling approaches in public health, land cover change modelling, and uncertainty in GIScience. He is the author of Using Historical Maps in Scientific Studies Applications, Challenges, and Best Practices (2019). He is currently pursuing research on three overlapping thematic areas of computational GIScience, developing pattern recognition techniques for feature extraction in historical maps, exploring dynamic phenomena in complex integrated systems in space and time, and generating geospatial models under uncertainty using probabilistic approaches and fuzzy set theory.
Yao-yi Chiang is an Associate Professor (Research) of Spatial Sciences in the Spatial Sciences Institute and the Associate Director of the Data Science Institute at the University of Southern California, Viterbi School of Engineering. He is an expert in digital map processing, pattern recognition, and geospatial information systems (GIS), and predictive analytics focusing on automated techniques for geospatial data extraction and integration. His paper “Querying Historical Maps as a Unified, Structured, and Linked Spatiotemporal Source” (2015) won the first prize at the Computing Community Consortium.
Mr Jack Reed is a former geologist and Geospatial Web Engineer in Stanford University Libraries’ Digital Library Systems and Services group where he works on increasing access to digital library content. He works with various open source software projects such as GeoBlacklight, Blacklight, and Leaflet-IIIF.
Dr Katherine McDonough is a historian of eighteenth-century France working at the intersection of political culture and the history of science and technology. She is part of the Living with Machines project at the Alan Turing Institute, where her research focuses on developing methods for geographic information retrieval from text and visual sources and examining how the expansion of transportation infrastructure changed 19th century communities. She is also involved in Spatial History of the Encyclopédie, a project on the history of early modern geographic information in collaboration with the GéoDisco group and the Textual Optics Lab/ARTFL Project at the University of Chicago.
Assoc Prof Andrea Nanetti is Associate Professor at NTU and a pioneer in the digital humanities. His main research project is Engineering Historical Memory (EHM), which develops and tests sets of shared conceptualisations and formal specifications for content management systems at the intersections of humanities and data science. His latest article “Unlocking historical data with artificial intelligence” (2019) discusses EHM and focuses on the conservation of human experiences, interactive global histories, and heritage in Afro-Eurasia’s pre-modern history.