| Resource Description Framework (RDF) is a standard data model for describing Web resources. Since Web data are semi-structured in nature, an RDF query Language should provide a mechanism for querying semi-structured data. The optional matching in SPARQL (draft) which is an RDF query language and proposed by W3C recently is just such a query mechanism. This paper lays emphasis on the query processing technology for optional matching in SPARQL. After analyzing the syntax and semantics of SPARQL's optional matching and summarizing existing query processing method for optional matching and its deficiencies, this paper proposes an query processing algorithm which supports not only simple optional matching but also multiple and nested optional matching, as well as complex multiple nested optional matching, and analyzes time complexity of the algorithm in theory. On the basis of algorithm design and theoretic analysis, this paper designs and implements an prototype system ROMQS for optional matching query, which supports for user's submitting of SPARQL queries and RDF graphs to be queried, executing queries, browsing query results. For carrying out experimental validation, many RDF graphs with different size which contain incomplete and irregular information are designed, and a group of SPARQL optional matching queries are also designed against these RDF graphs. Through comparing ROMQS's executing results with theoretic right results, the validity of optional matching query processing algorithm is validated. Based on a mass of experiment data, time complexity of the algorithm is analyzed. Through comparing ROMQS with other RDF query systems, ROMQS's advantage of functionality, that is being able to process multiple and nested optional matching, is verified. |
PDF Full Text Request