Distributed dynamic semantic query and reasoning for graph data
Xingyu Peng, Ying Pan · 2021
With the massive increase of graph data such as RDF and knowledge graphs, how to effectively improve the query efficiency of massive graph data has become a research hotspot in graph data management. However, most of the existing graph query research has the problem of lack of support for dynamic data query, and usually requires a high upfront cost to semantically integrate the data to provide effective semantic query and reasoning. In response to the above problems, this paper proposes a distributed dynamic semantic query and reasoning framework for graph data. Firstly, an initial query is created according to the user's search requirements, and then based on the ontology, Wikipedia, and other knowledge bases, the initial query is expanded by semantic concepts and semantic associations to create an expanded query. In a distributed environment, the expanded query is decomposed into multiple sub-queries. Then, the relevant semantic inference rules are matched, and query processing is performed on the distributed data so that multiple sub-query results will be generated. Finally, we evaluate and sort these results, aggregate the results of sub-queries with high scores and return them to the user. The framework supports dynamic semantic query and reasoning in a distributed environment, and can dynamically optimize and adjust the extended query and semantic reasoning rules in time, so as to provide support for real-time graph data query.