基于多级索引的前向语义数据流推理

语义数据流推理需要对实时高速的RDF数据进行高效推理.针对目前语义数据流处理中推理效率低、查询延迟高等难题,提出一种基于多级索引的前向实时推理机制.基于规则依赖建立推理规则触发顺序;针对开销最大的传递规则进行优化;构建基于传递规则的结果集子图索引;建立基于变量位置的三元组索引消除中间结果;通过查询条件依赖关系进行连接操...

FORWARD SEMANTIC DATA STREAM REASONING BASED ON MULTI-LEVEL INDEX

Semantic data stream reasoning requires efficient reasoning for real-time and high-speed RDF data.In the current semantic data stream processing,there are problems of low reasoning efficiency and high query delay.We propose a forward real-time reasoning mechanism based on multi-level indexing.The inference rule triggering order was established based on the rule dependency.We optimized the most expensive transitive rule to construct the result set subgraph index based on the transitive rule.A triple index based on the variable position was established to eliminate the intermediate results.Through the query of condition dependency relationship,we performed the connection to get the set that met the conditions.This paper used the LUBM dataset to verify the method proposed in this paper.The experimental results show that the real-time reasoning method based on multi-level indexing has obvious advantages over the current real-time reasoning engine in terms of the query delay and the throughput.