Machines Reading Maps
Machines Reading Maps - Composite Masonry View
In collaboration with the David Rumsey Map Collection, Luna Imaging partnered with research teams from the Alan Turing Institute, the University of Minnesota, the Austrian Institute of Technology and the University of Southern California Library.
The results of this project empowers researchers with access to the text contained on the maps - not just the descriptive data captured by catalogers - but more obscure references that may have once remained hidden from search results.
3 technology collaborators combine efforts
MapKurator (machine learning)
Advanced text searching as developed by the University of Minnesota enables text recognition for search and discovery beyond what appears in catalogued descriptions. Individual maps are processed and text and location are identified for discovery in LUNA. An initial subset of 56,000 images in the David Rumsey Map Collection resulted in 89 million indexed terms geo referenced on the maps. Learn more…
LUNA (visualization)
Using the David Rumsey Map Collection for the MRM’s proof-of-concept, the team at Luna Imaging expanded search options to encompass the newly indexed text on the maps..
Results appear with grid and masonry display options to provide a visual confirmation of the search results.
Proximity searching is possible with an advanced search to perform multiple word searches using the spatial search feature in Solr for related words.
3 - Annotorious (interactivity)
Newly indexed text on the maps can be identified, edited, and enhanced through the addition of the Annotorious annotation toolset in the LUNA Detail View.
Benefits for digital humanities scholars and researchers
Read about the potential outcomes for MRM in the Journal of Cultural Analytics
Searching maps by words: how machine learning changes the way we explore map collections by Valeria Vitale PhD, Assistant Professor in Digital Humanities, University of Sheffield and Machine Reading Maps team memberThe development conducted for the MRM project will inform future releases of LUNA. The collaboration has provided the team with a roadmap for moving forward in the following areas:
Discovery
The MRM project introduces the capacity to search text on maps within the David Rumsey Map Collection.
Annotations -
Text discovered on the maps provide supplemental annotations in addition to the searchable public and private annotations (Open Annotations and the IIIF Content Search API).
Guide to Searching and Annotating Text on Maps
Benefits for the LUNA community
The development conducted for the MRM project will inform future releases of LUNA and provide our team an exciting direction for image annotations, translations, transcription, and geospatial discovery.
Searching Text on Maps
Watch this LUNA Learning Video for Searching Text on Maps