Sparklis

From Semantic Web Standards
Jump to: navigation, search

Sparklis

Name of the tool: Sparklis
Home page: http://www.irisa.fr/LIS/ferre/sparklis/
Date of latest release: 2014-12-11 (verify)
Programming language(s) that can be used with the tool: OCaml, Javascript
Relevant semantic web technologies: RDF, SPARQL
Categories: Browser, Special Browser
See also:
Public mailing list:
Preferred project URI:
DOAP reference:
Company or institution:

(Tool description last modified on 2015-01-9.)

Description

Sparklis is a Web application to explore and query SPARQL endpoints. It combines at the same time faceted search, interactive query builders, and natural language interfaces. It is not tied to a particular endpoint, and works with any endpoint provided that it grants public or local access.

The principle of Sparklis is to let users see and explore data and build expressive queries in natural language at the same time. The SPARQL query is only visible at the bottom of the page, and is only for curious expert users. Users don't need to know the RDFS schema, and discover it on the fly. They don't need to write anything, apart from filter values (e.g., matching keywords), which ensures that none of lexical, syntactic, and schema errors are introduced.

Sparklis covers a large fragment of SPARQL: graph patterns, optional, union, negation, ordering, aggregation, main filters (string matching, inequalities and intervals, language or datatype). For example, here is a complex query on DBpedia that can be reached in about 20 navigation steps.

   Give me a writer
      that is the writer of the highest-to-lowest number of film
      and whose label has a language matching en
      and that optionally has a depiction
      and that optionally has a list of birth date
      and that has any birth place that has country France

It returns a list of film writers born in France with their English name, an optional depiction, an optional list of birth dates, and their number of films (in decreasing order). You can directly access that query from the "Examples" page, and learn about its construction from the demo/tutorial video.