Linked Traces annotations

v0.2 Draft for comment, 19 October 2019.

Pursuant to suggestions, the scope of Linked Traces has been modified. In the first draft (v0.1) only annotations having place identifier content were considered, in order to meet requirements of the Peripleo and World Historical Gazetteer platforms. By consensus, Linked Traces should concern Web Annotations of historical entities more generally, and include or reference multiple domain- and application-specific patterns.

Traces

The term "trace" refers to web-published resources concerning historical entities of any kind, and conventionally, to the entities themselves. Trace data takes the form of a W3C web annotations. The annotation Target is a web resource of any type (Text, Image, Dataset, Video, Sound), and the annotation Body either embeds RDF as text in a "value" or "textBody" attribute, or links to an external RDF record with its "id" attribute. Annotations have motivations, which include describing, linking, classifying, and tagging. It is possible to create new motivations; e.g. transcribing seems broadly useful.

Traces have already been indexed and displayed in the Peripleo web application developed by Rainer Simon for the Pelagios project; to date these are principally records of coins and inscriptions annotated with identifiers for places as "findspots" in the Classical Mediterranean region.

In this new conception, Linked Traces amounts to a set of use patterns for the W3C Web Annotation model and vocabulary to accommodate

• more kinds of historical entities (not only artifacts, but events of all kinds, people, and works, broadly construed),
• more motivations and use scenarios, e.g. transcriptions of page images
• more detailed content within, or referenced by, the annotation Body

Issues

1. What are use scenarios for Linked Traces annotations?
2. What should be the model patterns for annotation Body RDF in each domain/scenario?
3. Are those patterns outside the scope of this work?
4. When should Body RDF be embedded as opposed to externally linked to?

Scenarios and examples

A preliminary list, to be expanded. NB: W3C Annotations are specified using JSON-LD format (an RDF syntax).

• The geographic case: Annotations of web resources about any sort of historical item with identifiers for relevant places.

The World Historical Gazetteer (WHG) and Peripleo platforms solicit trace annotation contributions, the bodies of which include RDF expressing a setting (a place IRI and optional 'when' statement) and its relation to the entity referenced by the Target (e.g. 'waypoint', or 'findspot'). WHG software will display (and return via API), for an indexed place, any contributed traces associated with it, exposing the Target IRI and the 'when' and 'related' properties embedded in the Body. In this way, traces provide enhanced place description for users of WHG, e.g. place1234 was a waypoint on these several journeys centuries apart, was the findspot for these several artifacts, and was among the settings for this complex historical event.

# places in the lifepath of Gautama Bhudda; json array in body.value

{ "@context":["http://www.w3.org/ns/anno.jsonld",
    { "lpo": "http://linkedpasts.org/ontology/lpo.jsonld"}],
  "id": "http://example.org/annotations/12345",
  "type": "Annotation",
  "creator": {
    "id":"https://orcid.org/0000-1234-2345-6789",
    "name":"Anne Annotator",
    "homepage":"http://annotationist.org"
  },
  "created": "2019-10-18",
  "motivation": "describing",
  "target": [{
    "id": "https://en.wikipedia.org/wiki/Gautama_Buddha",
    "type": "Text",
    "format": "text/html",
    "title": "Gautama_Buddha"
  }],
  "body": {
    "type": "Dataset",
    "format": "application/json",
    "value": [{ 
      "id": "http://whgazetteer.org/places/86438",
      "title": "Lumbini",
      "relation": "lpo:birthplace",
      "when": {"timespans":[{"start": "-0563/-0480"}]}
    },{ 
      "id": "http://whgazetteer.org/places/86001",
      "lpo:title": "Kusinagara",
      "lpo:relation": "lpo:deathplace",
      "lpo:when": {"timespans":[{"start": "-0483/-0400"}]}
    }]
  }
}

• Transcription case #1: Annotations of archival page images with structured transcriptions of them.

Free UK Genealogy is organizing transcription of archival birth registries, which which they will store as RDF and make available via an API. In this case, the page image is the annotation Target (with an IRI as "id"), and the Body could either be only another IRI, to the transcription record for the page, or it could embed that RDF as media type application/rdf in a "value" field.

Body type A (external):

RDF transcription of data from an image of a birth register page 

{
  "@context": "http://www.w3.org/ns/anno.jsonld",
  "id": "http://example.org/anno_0001",
  "type": "Annotation",
  "license”: "<https://opendatacommons.org/licenses/odbl/1-0/>”,
  "creator”: "https://freeukgenealogy.org/volunteers/richardofsussex",
  "created": "2017-05-24",
  "motivation”: "transcription”,
  "target": {
    "type": "Image",
    "id": "https://freeukgenealogy.org/data/b65432"
  },
  "body": {
    "type": "Dataset",
    "format": "application/rdf"
    "id”: "http://example.org/birthregistration/1234”
  }
}

Body type B: (embedded)

  "body": {
    "type": "Dataset",
    "format": "application/rdf"
    "value”: '
        @prefix fugont: <https://freeukgenealogy.org/ontology.rdf#> .
        @prefix freeukgen: <https://freeukgenealogy.org/data/> .
        @prefix fugregdist: <https://freeukgenealogy.org/registrationdistricts/> .
        @prefix fugreg: <https://freeukgenealogy.org/registers/> .
        @prefix dc: <http://purl.org/dc/elements/1.1/> .
        @prefix dcterms: <http://dublincore.org/documents/2012/06/14/dcmi-terms/> .
        @prefix crm: <http://www.cidoc-crm.org/sites/default/files/cidoc_crm_v6.2-draft-2015August.rdfs#> .
        freeukgen:bmd12345 a fugont:transcription;
        	crm:P70_documents freeukgen:be23456;
        	dc:creator fugvol:richardofsussex;
        	dc:date "2017-05-24"^^xsd:date;
        	dcterms:license <https://opendatacommons.org/licenses/odbl/1-0/> . 
        freeukgen:be23456 a fugont:registrationevent;
        	crm:P70_documents  freeukgen:b65432;
        	crm:P7_took_place_at fugregdist:Droxford;
        	crm:P70i_is_documented_in fugreg:Droxford_7_72;
        	crm:P4_has_time_span [
        		crm:P82a_begin_of_the_begin "1850-07-01T00:00:00"^^xsd:dateTime; 
        		crm:P82b_end_of_the_end "1850-09-30T23:59:59"^^xsd:dateTime .
        	] .
        freeukgen:b65432 a crm:E67_Birth;
        	crm:P98_brought_into_life freeukgen:b65432_child;
        	crm:P96_by_mother freeukgen:b65432_mother;
        	crm:P97_from_father freeukgen:b65432_father;
        	crm:P4_has_time_span [
        		crm:P82a_begin_of_the_begin "1849-07-01T00:00:00"^^xsd:dateTime; 
        		crm:P82b_end_of_the_end "1850-09-30T23:59:59"^^xsd:dateTime .
        	] .
        freeukgen:b65432_child a crm:E21_Person;
        	crm:P1_is_identified_by "Light, Thomas Edward" .   
    '
  }

• Transcription case #2: Annotations of ___ with ___.

Recogito is an online tool for geo-tagging of text and image documents. In Recogito's case, the traces are the annotations that link the geographical references in the document to places in the gazetteer. Annotation bodies are simple, in this case, and point to the gazetteer URIs.

Example 1: Annotations on text

{
  "@context" : "http://www.w3.org/ns/anno.jsonld",
  "id" : "https://recogito.pelagios.org/annotation/533fb599-9e02-4fe2-ae98-6857b6055c22",
  "type" : "Annotation",
  "generator" : {
    "id" : "https://recogito.pelagios.org/",
    "type" : "Software",
    "name" : "Recogito",
    "homepage" : "https://recogito.pelagios.org/"
  },
  "generated" : "2019-10-21T10:47:56+00:00",
  "body" : [ {
    "type" : "SpecificResource",
    "value" : "http://pleiades.stoa.org/places/530906",
    "creator" : "https://recogito.pelagios.org/rainer",
    "modified" : "2019-10-21T10:47:52+00:00",
    "purpose" : "identifying"
  } ],
  "target" : {
    "source" : "https://recogito.pelagios.org/part/5ec8253d-f398-4355-82d8-ba7f324ea935",
    "type" : "Text",
    "selector" : [ {
      "type" : "TextQuoteSelector",
      "exact" : "Ithaca"
    }, {
      "type" : "RangeSelector",
      "startSelector" : {
        "type" : "XPathSelector",
        "value" : "/TEI[1]/text[1]/body[1]/div[1]/p[2]"
      },
      "endSelector" : {
        "type" : "XPathSelector",
        "value" : "/TEI[1]/text[1]/body[1]/div[1]/p[2]"
      }
    } ]
  }
}

Example 2: Annotations on a digitized old map

{
  "@context" : "http://www.w3.org/ns/anno.jsonld",
  "id" : "https://recogito.pelagios.org/annotation/041053ec-46f1-4cd9-89cd-1971926b3f97",
  "type" : "Annotation",
  "generator" : {
    "id" : "https://recogito.pelagios.org/",
    "type" : "Software",
    "name" : "Recogito",
    "homepage" : "https://recogito.pelagios.org/"
  },
  "generated" : "2019-10-21T10:46:18+00:00",
  "body" : [ {
    "type" : "TextualBody",
    "value" : "Helgapelt",
    "creator" : "https://recogito.pelagios.org/rainer",
    "modified" : "2019-10-21T10:46:10+00:00",
    "purpose" : "transcribing"
  }, {
    "type" : "SpecificResource",
    "value" : "http://sws.geonames.org/3412888",
    "creator" : "https://recogito.pelagios.org/rainer",
    "modified" : "2019-10-21T10:46:10+00:00",
    "purpose" : "identifying"
  } ],
  "target" : {
    "source" : "https://recogito.pelagios.org/part/46c9126b-6904-4229-bfc3-06b40d1834f1",
    "type" : "Image",
    "selector" : [ {
      "type" : "FragmentSelector",
      "conformsTo" : "http://www.w3.org/TR/media-frags/",
      "value" : "xywh=pixel:1992,3092,274,274"
    } ]
  }
}

---

Contributors: Karl Grossner (t,gh: @kgeographer); Rainer Simon (t: @aboutgeo, gh:@rsimon), Richard Light (t: @RichardOfSussex, Johannes Scholtz (t:@Joe_GISc)