Linked Data for Pleiades Places

Linking open data

Linked Data is an approach to integrating information across organizational and disciplinary boundaries characterized by the following principles:

1. Use URIs as names for things.
2. Use HTTP URIs, so that people can look up those names.
3. When someone looks up a URI, provide useful information, using the standards (RDF, SPARQL).
4. Include links to other URIs, so that they can discover more things.

"Open" in this context means licensed so that information about things can used without having to negotiate for access and permission. The almost 300 datasets adhering to these principles are shown in the diagram below (click it for a more link-rich expression of the diagram).

Pleiades has accumulated enough links to the DBpedia set that it qualifies for inclusion in the diagram. Other ancient world datasets that link to Pleiades can then tag along (or qualify now by linking to other included datasets). The individual datasets, and the digital arms of the research community as a whole, benefit from raising their profiles in this way.

The Big RDF Dataset

The Pleiades places dataset consists of more than 210,000 subject-predicate-object statements (or "triples") about more than 34,000 ancient places. It is written to a single text file using the terse Turtle RDF (resource description framework) format. Triples are expressed in Turtle like this:

# Not a complete Turtle document
:subject :predicate :object .

Or in the case of multiple statements with the same subject:

# Not a complete Turtle document
:subject :predicate1 :object1;
         :predicate2 :object2 .

The data specifically about Athens is shown below.

@prefix foaf: <http://xmlns.com/foaf/0.1/> .
@prefix geo: <http://www.w3.org/2003/01/geo/wgs84_pos#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix skos: <http://www.w3.org/2004/02/skos/core#> .
@prefix spatial: <http://geovocab.org/spatial#> .

<http://pleiades.stoa.org/places/579885#this> a <http://geovocab.org/spatial#Feature>;
    rdfs:label "Athenae";
    spatial:C <http://pleiades.stoa.org/places/579888#this>;
    rdfs:comment "A major Greek city-state";
    rdfs:seeAlso <http://dbpedia.org/resource/Athens>;
    geo:lat 37.970793;
    geo:long 23.719537;
    skos:altLabel "Athenae",
        "Athens";
    foaf:primaryTopicOf <http://pleiades.stoa.org/places/579885> .

The title and description of the Athens resource are carried by rdfs:label and rdfs:comment properties while attested forms of its names or their transliterations are carried by skos:altLabel properties. A single point representing the location of Athens is carried by geo:lat and geo:long properties. A link to the Pleiades web page on Athens, via which all of these triples are editable, is carried by the foaf:primaryTopicOf property and a link to the DBpedia resource on Athens (are thereby Wikipedia) is made with an rdfs:seeAlso. Those four namespaces (prefixed foaf, geo, rdfs, and skos) are among the most widely used in linked open data and ensure a good baseline for usability.

The namespace prefixed spatial is new and less well known. It's used to convey an assertion that Athens is geographically connected with Attica (identified by the URI http://pleiades.stoa.org/places/579888#this). There is only one such connection asserted presently, but It's important to keep in mind that Pleiades is an open world: no expressed connection to the Aegean Sea (for example) does not mean that Pleiades has ruled out that connection.

It takes about an hour to write the big dataset and we'll do so every other week at the most.

RDF Views of Individual Places

Data about individual places, on the other hand, will be kept completely current. For Athens see http://pleiades.stoa.org/places/579885/turtle (text/turtle) or http://pleiades.stoa.org/places/579885/rdf (application/rdf+xml) and note that any linked data client will be properly redirected to the appropriate format variant when dereferencing the identifier http://pleiades.stoa.org/places/579885#this that appears in the RDF data. The "#this" on the end of that URI, by the way, lets us distinguish between the ancient place itself which we cannot access via a computer network and the web page describing it which can be fetched from http://pleiades.stoa.org/places/579885.

Resources

The Linked Data book by Tom Heath and Christian Bizer was enormously helpful and an essential reference for the terminology used in this blog post. The rdflib package for Python is what our pleiades.rdf package uses to build the graphs, and does all the heavy lifting.

Postscript

On 2011-12-05 the Pleiades RDF dataset was added to the LOD Cloud and will appear in the next edition of the diagram.