Get instances of RDF triple patterns from LOD Laundromat. For a given RDF triple pattern, the tools looks up datasets that might have its instances via LOD Laundromat Index, then retrieves the instances from these datasets exposed via Triple Pattern Fragments endpoints.
Does basically some of what Frank does.
You can use the tool via its command-line interface. Either download a released executable of the tool or compile it yourself via Leiningen by running lein bin. In order to run the tool, you need to have Java (version 1.6 or newer) installed.
The pattern_launder executable reads RDF triple patterns from the standard input stream and writes its results to the standard output stream. This makes it possible to use it in data processing pipelines. The input triple patterns are provided in CSV format with the columns subject, predicate, and object. The values of these columns are either RDF terms, serialized according to the Triple Pattern Fragments specification, or empty, indicating a variable. For example, the following are some triple patterns for SKOS:
subject,predicate,object
,http://www.w3.org/2004/02/skos/core#relatedMatch,
,http://www.w3.org/2004/02/skos/core#closeMatch,
,http://www.w3.org/2004/02/skos/core#narrowMatch,
By default, the tool retrieves data matching each triple pattern and outputs it in the N-Triples RDF syntax. Alternatively, you can ask for estimated counts of triples matching given triple patterns by using the --counts command-line switch. In that case, a count column with the estimated counts will be added to the input CSV with triple patterns, like this:
subject,predicate,object,count
,http://www.w3.org/2004/02/skos/core#relatedMatch,,118806
,http://www.w3.org/2004/02/skos/core#closeMatch,,298452
,http://www.w3.org/2004/02/skos/core#narrowMatch,,2738
You can see all the command-line options of the tool by running ./pattern_launder --help.
For a more detailed example, you can see how to estimate popularity of vocabularies.
Alternatively, you can use the tool as a dependency in your Clojure programs. The pattern-launder.core namespace exposes two public functions matching the command-line functionality. The function triples obtains triples matching a given triple pattern (a map with :subject, :predicate, and :object keys) and returns a lazy sequence of triples in JSON-LD. The function triple-count returns an estimated count of triples matching the given triple pattern, provided in the same format as above.
Copyright © 2018 Jindřich Mynarz
Distributed under the Eclipse Public License version 1.0.