search-lucene

Clojure-based software for text searching and mapping terms with IRIs.

Mapping file

Mapping file consists two blocks:

• mapping terms with subjects. Single subject for single term

Example:

@prefix map: <http://rdf.adalab-project.org/ontology/mapping/> .
@prefix skos: <http://www.w3.org/2004/02/skos/core#> .

map:d586122a836de73dec223eb0780d25f40bb51269 a map:MappedTerm ;
   map:sha1 "d586122a836de73dec223eb0780d25f40bb51269" ;
   skos:notation "term1" ;
   skos:closeMatch <http://example.org/data#record1> .

• declare weights for properties

Example:

 @prefix map: <http://rdf.adalab-project.org/ontology/mapping/> .

 :current_weights a map:WeightSet ;
   map:weights [ rdf:predicate :term ; map:weight "1.5"^^xsd:float ] ,
               [ rdf:predicate :term2 ; map:weight "0.5"^^xsd:float ] ,
               [ rdf:predicate :term50 ; map:weight "3"^^xsd:float ] .

The text searching feature is delivered by Apache Lucene framework. Location of the repository with the lucene index is given by option --repository-dir.

Options

• -h, --help Print help screen
• -t, --terms-file FILE File containing terms to search. Single term in line.
• -m, --mapping-file FILE File containing mapping data including wages.
• -r, --repository-dir DIR Location where the native repository should be put.

Usage

java -jar search-lucene-standalone.jar -t terms.csv -m mapping_file.ttl -r /tmp/repository