Abstract:
The task of accessing knowledge graphs through structured query languages like SPARQL is rather demanding
for ordinary users. Consequently, various approaches exploit the simpler and widely used keyword-based search paradigm, either by translating keyword queries to structured queries, or by adopting classical information retrieval (IR) techniques. In this paper, we study and adapt Elasticsearch, an out-of-the-box document-centric IR system, for supporting keyword search over arbitrary RDF datasets. Contrary to other works that mainly retrieve entities, we opt for retrieving triples, due to their expressiveness and informativeness. We specify the set of functional requirements and study the emerging questions related to the selection and weighting of the triple data to index, and the structuring and ranking of the retrieved results. Then we perform an extensive evaluation of the different factors that affect the IR performance for four different query types. The reported results are very promising and offer useful insights on how different Elasticsearch configurations affect the retrieval effectiveness and efficiency.
Overall, the proposed method is a scalable and efficient method for keyword search over RDF with effectiveness comparable to the effectiveness of dedicated task- and dataset-specific systems (as evaluated using DBpedia-Entity test collection for entity search). It is triple-centered, enabling the provision of more precise and explainable results, and relies on a special configuration of Elasticsearch for the needs of RDF that is schema agnostic, and thus widely applicable. Finally we briefly describe Elas4RDF, a system that is powered by the proposed approach.
