A fuzzy semantic representation and reasoning model for multiple associative predicates in knowledge graph
Pu Li,Xin Wang,Hui Liang,Suzhi Zhang,Yazhou Zhang,Yuncheng Jiang,Yong Tang
DOI: https://doi.org/10.1016/j.ins.2022.03.079
IF: 8.1
2022-06-01
Information Sciences
Abstract:As the latest achievement of the development in semiotics, knowledge graph has been recognized and widely used by more and more researchers for its rich semantic information and clear logical structure. How to discovery the deep relevant knowledge from the massive graph-structured data has become a hot spot of artificial intelligence. Considering that some predicates in knowledge graph express fuzzy relationships whose semantics are not certain, the basic schema of classical knowledge graph in the form of RDF triple cannot describe the fuzzy semantic information effectively. To counter above problems, in this paper, we present a new semantic representation and reasoning model for multiple associative predicates by introducing fuzzy theory. Concretely, the presented method defines a new fuzzy annotating strategy to represent the fuzzy semantics between associative predicates in different RDF triples. On this basis, some fuzzy reasoning rules are presented to realize fuzzy semantic extension for classical knowledge graph. Lastly, the experimental results show that our proposal can discover more implicit valid knowledge with fuzzy semantic and have a good consistency with the intuition of human judgments. Overall, the methods proposed in this paper constitute some effective ways of knowledge discovery of structured semantic data.
computer science, information systems