ddw-standard-index.ttl

@prefix vg:   <https://data.world/vg/1.0/> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix owl:  <http://www.w3.org/2002/07/owl#> .
@prefix aice: <https://data.world/aice/1.0/> .

### SIMPLE INDEXES (ALL CLASSES/PREDICATES IN DATASET)
# Index all classes and predicates in a dataset

aice:ClassIndex
    a               vg:VectorIndex ;
    aice:indexType  owl:Class ;
    vg:model        vg:mxbai_embed_large_v1 ;
    vg:indexerQuery """
    PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
    PREFIX dct:  <http://purl.org/dc/terms/>
    PREFIX aice: <https://data.world/aice/1.0/>

    SELECT ?iri (GROUP_CONCAT(?content; SEPARATOR = " ") AS ?value)
    WHERE {
        # Grab a bunch of predicates and concat them
        VALUES ?predicate { rdfs:label dct:title rdfs:comment dct:description }

        # pulls dynamically from the graph to allow for extension
        aice:ClassIndex aice:indexType ?class .

        ?iri a/rdfs:subClassOf* ?class.
        ?iri ?predicate         ?content.
    }
    GROUP BY ?iri
    """
.

aice:ObjectPropertyIndex
    a               vg:VectorIndex ;
    aice:indexType  owl:ObjectProperty ;
    vg:model        vg:mxbai_embed_large_v1 ;
    vg:indexerQuery """
    PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
    PREFIX dct:  <http://purl.org/dc/terms/>
    PREFIX aice: <https://data.world/aice/1.0/>

    SELECT ?iri (GROUP_CONCAT(?content; SEPARATOR = " ") AS ?value)
    WHERE {
        # Grab a bunch of predicates and concat them
        VALUES ?predicate { rdfs:label dct:title rdfs:comment dct:description }

        # pulls dynamically from the graph to allow for extension
        aice:ObjectPropertyIndex aice:indexType ?class .

        ?iri a/rdfs:subClassOf* ?class.
        ?iri ?predicate         ?content.
    }
    GROUP BY ?iri
    """
.

aice:DatatypePropertyIndex
    a               vg:VectorIndex ;
    aice:indexType  owl:DatatypeProperty ;
    vg:model        vg:mxbai_embed_large_v1 ;
    vg:indexerQuery """
    PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
    PREFIX dct:  <http://purl.org/dc/terms/>
    PREFIX aice: <https://data.world/aice/1.0/>

    SELECT ?iri (GROUP_CONCAT(?content; SEPARATOR = " ") AS ?value)
    WHERE {
        # Grab a bunch of predicates and concat them
        VALUES ?predicate { rdfs:label dct:title rdfs:comment dct:description }

        # pulls dynamically from the graph to allow for extension
        aice:DatatypePropertyIndex aice:indexType ?class .

        ?iri a/rdfs:subClassOf* ?class.
        ?iri ?predicate         ?content.
    }
    GROUP BY ?iri
    """
.

### MAPPED INDEXES (ONLY MAPPED CLASSES/PREDICATES IN DATASET)
# These indexes pull from the r2rml to determine the correct classes/predicates to index

aice:MappedClassIndex
    a               vg:VectorIndex ;
    vg:model        vg:mxbai_embed_large_v1 ;
    vg:indexerQuery """
    PREFIX rr: <http://www.w3.org/ns/r2rml#>
    PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>

    SELECT ?iri (GROUP_CONCAT(?content; SEPARATOR = " ") AS ?value)
    WHERE {
        [rr:class ?class].
        ?iri rdfs:subClassOf*  ?class ;
             a owl:Class.
        VALUES ?contentPredicate {rdfs:label dct:title rdfs:comment dct:description}
        ?iri ?contentPredicate ?content.
    }
    GROUP BY ?iri
    """
.

aice:MappedObjectPropertyIndex
    a               vg:VectorIndex ;
    vg:model        vg:mxbai_embed_large_v1 ;
    vg:indexerQuery """
    PREFIX rr: <http://www.w3.org/ns/r2rml#>
    PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
    PREFIX owl: <http://www.w3.org/2002/07/owl#>

    SELECT ?iri (GROUP_CONCAT(?content; SEPARATOR = " ") AS ?value)
    WHERE {
        [rr:predicate ?predicate].
        ?iri       rdfs:subPropertyOf* ?predicate.
        ?predicate a                   owl:ObjectProperty.
        VALUES ?contentPredicate {rdfs:label dct:title rdfs:comment dct:description}
        ?iri       ?contentPredicate   ?content.
    }
    GROUP BY ?iri
    """
.

aice:MappedDatatypePropertyIndex
    a               vg:VectorIndex ;
    vg:model        vg:mxbai_embed_large_v1 ;
    vg:indexerQuery """
    PREFIX rr: <http://www.w3.org/ns/r2rml#>
    PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
    PREFIX owl: <http://www.w3.org/2002/07/owl#>
    PREFIX csvw: <http://www.w3.org/ns/csvw#>

    SELECT ?iri (GROUP_CONCAT(?content; SEPARATOR = " ") AS ?value)
    WHERE {
        [rr:predicate ?predicate].
        ?iri       rdfs:subPropertyOf* ?predicate.
        ?predicate a                   owl:DatatypeProperty.
        FILTER NOT EXISTS { ?iri a csvw:Column }
        VALUES ?contentPredicate {rdfs:label dct:title rdfs:comment dct:description}
        ?iri       ?contentPredicate   ?content.
    }
    GROUP BY ?iri
    """
.