Introduction to the LCT-01 Holotype Standard
The LCT-01 Holotype Standard (LHS) is an RDF ontology for the representation of LCT-01 holotype data. Its development is in line with our desire for FAIR (findable, accessible, interoperable, and reusable) data. The largest source of LHS data so far is available through the LCT-01 Triplestore.
In addition to this document we also maintain an OWLDoc at https://lct01.lemuria.ph/rdf/owldoc/lct01/.
Data model
In brief, the most important parts of the LHS are the taxon, image, and event.
Taxa
The taxon (plural taxa) is the basic building block of the LCT-01. It is a group of clothing, or of other taxa, treated as a single cohesive unit for classification purposes.
Images
One of the most distinct features of the LCT-01 Holotype Standard is the separation of images (ImageType) from where they are hosted (Image). An ImageType describes what happened in the image, who took it, what it looks like (through cryptographic hashes), and where it is. Describing “where it is” is done by pointing to Images (lhs:Image). Images contain an external URL linking to an image, information about the provider, and point back to the ImageType they describe.
The ImageType itself has cryptographic hashes for the image it represents, making sure one can always verify that the raw data at a URL an Image points to actually was the raw data inspected when the holotype was designated. We use the cryptographic SHA2-256 and BLAKE3 algorithms for this.
Hashing images also provides storage and copyright benefits. Because we do not store or rehost the original images, but rather cite, hash, and link to them, we do not fall afoul of copyright law. This is our formally named principle of “Cite, Link, Hash, but don’t Embed” (and other variations on wording, but the core idea is the same.) But more importantly, we save on bandwidth and storage space.
Events
The LHS uses a system of events that indicate what happened in a picture; such as a WearEvent, VisibleGarmentEvent, HangingGarmentEvent, and a FlatlayEvent. There will be many more events to come and when the traditional data model doesn’t fit, it can simply be extended.