VesselAI Linked Open Data Set

The following example diagram illustrates the status of a specific vessel and for a given time instant:

The corresponding SPARQL query for the above diagram is:



xxxxxxxxxx
32
1
prefix geof:    <http://www.opengis.net/def/function/geosparql/> 
2
 prefix rdf:     <http://www.w3.org/1999/02/22-rdf-syntax-ns#> 
3
 prefix strdf:   <http://strdf.di.uoa.gr/ontology#> 
4
 prefix xsd:     <http://www.w3.org/2001/XMLSchema#> 
5
 prefix ogc:     <http://www.opengis.net/ont/geosparql#> 
6
 prefix rdfs:    <http://www.w3.org/2000/01/rdf-schema#> 
7
 prefix dul:    <http://www.ontologydesignpatterns.org/ont/dul/DUL.owl#> 
8
 prefix :     <http://www.vesselAI-project.eu/ontology#> 
9
 prefix wunit:     <http://www.vesselAI-project.eu/weatherUnits#> 
10
 prefix time: <http://www.w3.org/2006/time#> 
11
 PREFIX ex: <http://www.topology.org/>
12
 PREFIX s: <http://www.spatial.org/>
13
 PREFIX t: <http://www.temporal.org/>
14
 
15
SELECT DISTINCT * WHERE {
16
   [] :hasName "VERDI" ;
17
      :hasTrajectory ?tr .
18
   ?tr :hasPart ?node .
19
   ?node :hasGeometry ?geom ; 
20
      :hasTemporalFeature ?tf ;
21
      :hasWeatherCondition ?weather .
22
   ?weather :10_metre_U_wind_component_surface ?uwind ;
23
      :10_metre_V_wind_component_surface ?vwind ;
24
      :2_metre_dewpoint_temperature_surface ?dewpoint ;
25
      :2_metre_temperature_surface ?temp2m ;
26
      :Mean_sea_level_pressure_surface ?slp ;
27
      :Mean_wave_direction_msl ?waveDirection ;
28
      :Mean_wave_period_msl ?weavePeriod ;
29
      :Sea_surface_temperature_surface ?temp ;
30
      :Total_precipitation_surface_1_Hour_Accumulation ?precipitation .
31
   ?tf :hasTimeStart "2021-01-01 00:13:46" .
32
}

Often, it is necessary to combine semantic trajectories with contextual information available in the domain. For example, retrieving the list of vessels that have departed from seaports having certain facilities (e.g. passenger facilities).

The following listing implements the above example in SPARQL:


1
prefix geof: <http://www.opengis.net/def/function/geosparql/>
2
prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
3
prefix strdf: <http://strdf.di.uoa.gr/ontology#>
4
prefix xsd: <http://www.w3.org/2001/XMLSchema#>
5
prefix ogc: <http://www.opengis.net/ont/geosparql#>
6
prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
7
prefix :     <http://www.vesselAI-project.eu/ontology#>
8
prefix dul: <http://www.ontologydesignpatterns.org/ont/dul/DUL.owl#>
9
prefix wunit: <https://www.vessel-ai.eu/weatherUnit#>
10
PREFIX ex: <http://www.topology.org/>
11
PREFIX s: <http://www.spatial.org/>
12
13
SELECT DISTINCT ?vesselName ?vesselType ?name ?time WHERE {
14
   [] a ?vesselType ; :hasName ?vesselName ;
15
      :hasTrajectory ?tr .
16
   ?tr :hasStart ?start . 
17
   ?start :hasGeometry ?depGeom ; 
18
      :hasTemporalFeature ?tf . 
19
   ?tf :hasTimeStart ?time .
20
   ?port a :Port ; :hasName ?name ; :hasGeometry ?depGeom ; :hasPortFeatures ?features .
21
   FILTER(("2021-01-01 00:00:00"<?time) && (?time<"2021-01-05 00:00:00") && contains(lcase(str(?name)),"norway") && contains(lcase(str(?features)),"passenger") && (?vesselType!=:Vessel))
22
} ORDER BY ?time LIMIT 100

The spatial information contained in semantic trajectories can be exploited using spatial functions. However, since SPARQL 1.1 does not inherently suppport such functions, they are implemented and registered on the triple store using the namespace:

 PREFIX s: <http://www.spatial.org/>

For example, a query which retrieves the status of a specific vessel at a given time instant, and the nearest port with ship repair facilities, can be described in SPARQL 1.1 and the spatial distance function (registered on the triple store) s:distance() which returns the distance in Km between two geometries, as follows:


1
prefix geof: <http://www.opengis.net/def/function/geosparql/>
2
prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
3
prefix strdf: <http://strdf.di.uoa.gr/ontology#>
4
prefix xsd: <http://www.w3.org/2001/XMLSchema#>
5
prefix ogc: <http://www.opengis.net/ont/geosparql#>
6
prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
7
prefix :     <http://www.vesselAI-project.eu/ontology#>
8
prefix dul: <http://www.ontologydesignpatterns.org/ont/dul/DUL.owl#>
9
prefix wunit: <https://www.vessel-ai.eu/weatherUnit#>
10
PREFIX ex: <http://www.topology.org/>
11
PREFIX s: <http://www.spatial.org/>
12
13
SELECT DISTINCT ?portName ?vesselPos ?portPos ?dist WHERE { 
14
  {
15
  SELECT ?pname ?portPos ?features WHERE{   
16
    ?port a :Port ; :hasName ?pname ;
17
      :hasDescription ?features ;
18
      :hasGeometry ?pg .
19
    ?pg ogc:asWKT ?portPos .
20
    FILTER(
21
      contains(lcase(str(?features)),"ship repair") && contains(lcase(str(?pname)),"norway") ) .
22
    }
23
  }   
24
25
  {SELECT ?vesselPos ?t WHERE{
26
    :ves259490000_2021 :hasTrajectory ?tr .
27
    ?tr :hasPart ?node .
28
    ?node :hasGeometry ?g ; :hasTemporalFeature ?tf .
29
    ?tf :hasTimeStart "2021-01-11 16:16:28" .
30
    ?g ogc:asWKT ?vesselPos .
31
    }
32
  }
33
  BIND(s:distance(?portPos,?vesselPos) as ?dist) .
34
} 
35
ORDER BY ?dist 
36
LIMIT 1

Another query example, retrieves the reported positions found within a protected area, e.g. "Gule Rev". The topological function ex:within() has been registered on the triple store using the name space:

 PREFIX ex: <http://www.topology.org/>

The SPARQL 1.1 listing for the above example is as follows:


x
 
1
prefix geof: <http://www.opengis.net/def/function/geosparql/>
2
prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
3
prefix strdf: <http://strdf.di.uoa.gr/ontology#>
4
prefix xsd: <http://www.w3.org/2001/XMLSchema#>
5
prefix ogc: <http://www.opengis.net/ont/geosparql#>
6
prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>
7
prefix :     <http://www.vesselAI-project.eu/ontology#>
8
prefix dul: <http://www.ontologydesignpatterns.org/ont/dul/DUL.owl#>
9
prefix wunit: <https://www.vessel-ai.eu/weatherUnit#>
10
PREFIX ex: <http://www.topology.org/>
11
PREFIX s: <http://www.spatial.org/>
12
13
SELECT DISTINCT ?vname ?g ?pwkt WHERE { 
14
{
15
SELECT ?pwkt WHERE {
16
   ?p a :ProtectedArea ;
17
      :hasName "Gule Rev" ; :hasGeometry ?pg .
18
   ?pg ogc:asWKT ?pwkt .
19
}
20
}
21
22
23
{SELECT ?vname ?g WHERE {
24
   ?v :hasName ?vname ; a :FishingVessel ; 
25
      :hasTrajectory ?tr.
26
   ?tr :hasPart ?node .
27
   ?node :hasGeometry ?geom .
28
   ?geom ogc:asWKT ?g .
29
}
30
}
31
   FILTER(ex:within(?g, ?pwkt)) .
32
} LIMIT 100

The result sets of the above examples can be also rendered on 2D maps (using Geo tab below).

SPARQL 1.1 allows the description of federated queries, where the result set is computed from data available in two or more sources. In this example we assume that the information about marinas is available on a remote endpoint (WikiData.org) and we wish to combine it with information available locally (position of a vessel at a specific time instant). The following query, requests all the marinas known to be within a distance of 100Km from vessel named "VERDI" at "2021-01-01 00:13:46":


x
 
1
 PREFIX : <http://www.vesselAI-project.eu/ontology#>
2
PREFIX s: <http://www.spatial.org/>
3
PREFIX wdt: <http://www.wikidata.org/prop/direct/>
4
PREFIX wd: <http://www.wikidata.org/entity/>
5
 PREFIX geo: <http://www.opengis.net/ont/geosparql#>
6
7
 SELECT ?place ?location ?dist WHERE{
8
  {[] :hasName "VERDI" ; 
9
      :hasTrajectory/:hasPart ?node . 
10
  ?node :hasGeometry/geo:asWKT ?pos ; :hasTemporalFeature/:hasTimeStart "2021-01-01 00:13:46" .
11
  }
12
  SERVICE <https://query.wikidata.org/sparql> {
13
      SELECT ?place ?location WHERE {
14
         ?place wdt:P625 ?location ; wdt:P31 wd:Q721207 .
15
      }
16
  }
17
  BIND(s:distance(?pos,?location) as ?dist) .
18
  FILTER(?dist<100)
19
}

The result set of the above example can be also rendered on 2D maps (using Geo tab below).

VesselAI Linked Open Data Set Endpoint (testbed)