Big Historical Data conference - communication

[zoometh]

"Shared heritage": management and integration of cultural heritage data across Arches-based platforms in the Global South

Session: Environments of big cultural heritage data integration, 22-25 Nov 2023, Jena, Germany, Max Planck Institute of Geoanthropology, https://bhdc.earth/

Arches is a FOSS geospatial semantic web-based purpose-built platform for cultural heritage management. Several university projects on the heritage of the Global South are using Arches as an information system to assess the cultural heritage of large continuous geographical areas (sub-Saharan Africa, North Africa and the Middle East, Eurasian steppes, Amazon-Andes, etc.). The use of this software, in such contexts, tackles many of the challenges posed by digital humanities: large datasets, open access and open data, capacity building, digital gap, etc. Interoperability and sustainability are among the main challenges faced by Arches-based projects. To meet these challenges, we develop computer routines, in the form of computer scripts, and host them with reference data on a collaboration and version control platform (https://github.com/achp-project). Among these shared reference data sets, the so-called Arches Resource Models are CIDOC-CRM-compliant classes for describing and organising information related to cultural heritage in Arches: Places, Buildings, Geoarchaeological corings, etc. These graph structures can be parsed and compared with standardised metrics, such as the number of common subgraphs, to understand the similarities and differences in the assessments of cultural heritage made by the different Arches-based projects. We will present this FAIR approach to cultural heritage management, and discuss current and future challenges.

Authors and affiliations:

• Thomas Huet, University of Oxford, UK thomas.huet@arch.ox.ac.uk
• Alan Tapscott, Universitat Pompeu Fabra, Spain alan.tapscott@upf.edu
• Jonas Gregorio de Souza, Universitat Pompeu Fabra, Spain jonas.gregorio@upf.edu
• Junaid Abdul Jabbar, McDonald Institute for Archaeological Research, University of Cambridge, UK ja738@cam.ac.uk
• Manoj Lokare, McDonald Institute for Archaeological Research, University of Cambridge, UK ml2098@cam.ac.uk
• Ash Smith, Maritime Endangered Archaeology, University of Southampton, UK ash.smith@soton.ac.uk
• Mike Fisher, Department of Archaeology, Max Planck Institute of Geoanthropology michael.fisher@shh.mpg.de
• Mahmoud Abdelrazek, Advanced Research Computing Centre, UCL, UK mahmoud.abdelrazek@ucl.ac.uk
• Bryan Alvey, Central Asian Archaeological Landscapes, Institute of Archaeology, UCL, UK bryan.alvey@blueyonder.co.uk
• Marco Nebbia, UCL Institute of Archaeology, UK m.nebbia@ucl.ac.uk
• Krisztina Baranyai, Maritime Asia Heritage Survey, Kyoto University, Japan krisztina.baranyai@cseas.kyoto-u.ac.jp

[zoometh]

I suggest that we use a Jupyter notebook for the support of our presentation, mirrored on Google Colab for interactivity, see this first interactive widget to select the two Resource models to compare

@atapscott and @ads04r what do you think?

[zoometh]

MyBinder, as suggest by @razekmh, can be another solution. But it doesn't seems to handle GitHub file properly: changes can be done in MyBinder but are not updated on the GitHub file. See the same interactive widget

[razekmh]

I think the behaviour is there by design. Binder provides executable environment of the Jupyter notebooks in the repository but does not seem to have a facility to allow users to edit the original Jupyter notebook from the target repository. What I thought would be a good idea is to add a badge similar to this one => within our repository which directs users to an environment running Jupyter. We can add a few modification to the notebook or the repository to optimise the process but it is very light edit anyway.

[zoometh]

TODO:

• [x] rename the file
• [ ] review the code
• [ ] review the code documentation
• [x] add geographical extents of current projects in a leaflet map
• plot functions: plot_all_G(), plot_all_nx(), etc.
  • [ ] add legends (colors)
  • [ ] allow users to plot CIDOC-CRM names (ex: 'E54') or custom project labels (ex: 'Depth/Elevation')
• sup. 2 RMs comparisons
  • [ ] check both results
  • [ ] work on a symmetrical matrix/square matrix to highlight distances/correlations on RMs pairwise comparisons
• [ ] understand how g.show_buttons(filter_=["physics"]) works (plot_all_pyvis_G())
• [ ] evaluate Google Colab vs Binder