Knowledge-driven SPARQL query construction method

Abstract

The invention discloses a knowledge-driven SPARQL query construction method, and belongs to the technical field of data query. The method comprises the following steps of obtaining an SPARQL query statement submitted by a user, and extracting an entity set and a predicate set in the SPARQL query statement; analyzing and judging respectively, entity numbers and predicate numbers in an entity set and a predicate set, classifying query statements into simple questions, fact questions and complex questions respectively, dividing the complex questions into display entity questions and implicit entity questions, constructing query processes for different categories of questions respectively, and corresponding query answers are obtained. According to different problems, a corresponding knowledgegraph is constructed, and correct SPARQL query statements are screened out. The search space of a traditional method is reduced, and the operation time is shortened. When a complex problem is processed, if the implicit entity does not appear, the implicit entity is limited through an intermediate product, so that a correct SPARQL query statement is found out, and a result is executed accurately and quickly.

Description

technical field [0001] The invention relates to the technical field of data query, in particular to a knowledge-driven SPARQL query construction method. Background technique [0002] In the prior art, the traditional data query component method is mainly to arrange and combine the entities and predicates in the results of entity links and relationship links by violence, and then filter the constructed SPARQL query statements through some simple filtering rules. Due to the characteristics of SPARQL syntax, each SPARQL is composed of several triples, and each triple has two entity placeholders and one predicate placeholder. It is an item to exhaustively enumerate all the combinations Things with a lot of work. For example, from the question "List TV shows with producer as Erik Bork and company is Dream Works Television?", the entity set E can be extracted: [0003] {<http: / / dbpedia.org / resource / Erik_Bork>, [0004] <http: / / dbpedia.org / resource / DreamWorks_Television...

AI Technical Summary

Problems solved by technology

[0010] (1) The search space is large, and entities and predicates need to be fully arranged. The search space will increase exponentially with the increase of the number of entities and predicates, which will consume huge storage space

[0011] (2) The running time is long, and the correct answer needs to be found in the full arrangement space of entities and predicates, which will take a long time and the user experience is very bad

Images