关键词: RDF图, 不等长编码, 谓语标签树, 多元语义查询

Abstract: Knowledge graphs are semantic networks to reveal the relationships between entities, which are often expressed in the form of Resource Description Framework (RDF). Faced with the explosive growth of information, the existing approaches have the problem of ignoring the diversified semantic query requirements. Therefore, considering the rich se-mantic information of RDF graphs, a diversity semantic query with distributed processing based on multilevel neighborhood predicate label tree encoding index on RDF graphs is proposed. To avoid wasting storage space and assist subsequent parallel queries, a frequency-based predicate encoding mapping strategy is designed to map the predicates represented by long strings to unique natural number. Then the RDF graph is partition, the vertices are classified according to their adjacent edge properties and the corresponding storage mode is designed. After that, a multi-level neighborhood predicate label tree encoding index is proposed to filter invalid vertices and edges. For diversity semantic query with known predicate, known subject/object and known mixture, the corresponding matching strategies are given. An optimal connection based on common vertex is proposed to reduce cartesian product number and the cost of connection. The experimental results show that the query time can be improved by 5 to 9 times by using the constructed index for pruning optimization compared with without preprocessing. And compared with FAST algorithm, the query efficiency is improved by 43% on average. So the proposed index and query strategy can effectively deal with diversity semantic query on large-scale RDF graphs.

Key words: RDF graph, variable length encoding, predicate label tree, diversity semantic query