Update stat props for sketch unique counts

Issue: RiverBench/RiverBench#91
RiverBench · Sep 11, 2024 · 613a033 · 613a033
1 parent c36149f
commit 613a033
Show file tree

Hide file tree

Showing 2 changed files with 19 additions and 3 deletions.
diff --git a/src/documentation.ttl b/src/documentation.ttl
@@ -453,7 +453,7 @@ rb:standardDeviation rbdoc:hasLabelOverride "St. dev."@en ;
 rb:sum rbdoc:hasDocWeight "2100"^^xsd:int .
 
 
-rb:uniqueCount rbdoc:hasLabelOverride "Unique (approx.)"@en ;
+rb:uniqueCount rbdoc:hasLabelOverride "Unique"@en ;
                rbdoc:hasDocWeight "2101"^^xsd:int .
 
 

diff --git a/src/metadata.ttl b/src/metadata.ttl
@@ -423,8 +423,24 @@ rb:sum rdf:type owl:DatatypeProperty ;
 rb:uniqueCount rdf:type owl:DatatypeProperty ;
                rdfs:subPropertyOf rb:statisticalProperty ;
                rdfs:range xsd:integer ;
-               rdfs:comment "Only used for count statistics. Indicates how many unique elements are in the entire dataset. The value is estimated with a Bloom filter and is accurate to ~1%."@en ;
-               rdfs:label "Unique count (estimated)"@en .
+               rdfs:comment "Only used for count statistics. Indicates how many unique elements are in the entire dataset. The value may be estimated using a HyperLogLog sketch. In that case, rb:uniqueCountLowerBound and rb:uniqueCountLowerBound properties are also set on the subject."@en ;
+               rdfs:label "Unique count"@en .
+
+
+###  https://w3id.org/riverbench/schema/metadata#uniqueCountLowerBound
+rb:uniqueCountLowerBound rdf:type owl:DatatypeProperty ;
+                         rdfs:subPropertyOf rb:statisticalProperty ;
+                         rdfs:range xsd:integer ;
+                         rdfs:comment "Lower bound estimate of how many unique elements are in the entire dataset. The estimate is given for a 95% confidence interval."@en ;
+                         rdfs:label "Unique count lower bound estimate"@en .
+
+
+###  https://w3id.org/riverbench/schema/metadata#uniqueCountUpperBound
+rb:uniqueCountUpperBound rdf:type owl:DatatypeProperty ;
+                         rdfs:subPropertyOf rb:statisticalProperty ;
+                         rdfs:range xsd:integer ;
+                         rdfs:comment "Upper bound estimate of how many unique elements are in the entire dataset. The estimate is given for a 95% confidence interval."@en ;
+                         rdfs:label "Unique count upper bound estimate"@en .
 
 
 ###  https://w3id.org/riverbench/schema/metadata#usesGeneralizedRdfDatasets