HCMC Journal

LINCS project integration: meeting with LINCS team

17 June 2022: Martin Holmes
Minutes: 90

Met with the LINCS team to go over key issues, with this outcome:

• Our aim is to produce the Turtle output required directly from the MoEML source, to avoid intermediate formats. LINCS will provide sample code we can use as a model.
• Placenames are datable; i.e. a place may have multiple names over a period of time.
• Locations are datable: i.e. a place may move.
• Functions (which MoEML plans to encode using <state>) are not datable; a change in function supposes a new place.
• We will have to convert our Julian dates to proleptic Gregorian, to place in xsd:date type attributes, but also provide the original Julian in a text node.
• We will include the list of alternative names for each place from our gazetteer, but we will not (in the first round of ingestion) provide sources or dates for those names. We may do that in a later round.
• Where we do not have dates, we will assume the MoEML date range to be the range for our records, and insert it.
• Action on MDH to create diagnostics which provide some insight into the number of places with complex scenarios, as well as a repair list of anomalous cases (two locations elements, but only one dated, for example).