面向语义网的本体不确定性推理建模仿真

Modeling and Simulation of Ontology Uncertainty Inference for Semantic Web

研究面向语义网的本体不确定性推理建模仿真方法,准确推理语义网内不确定性数据,为各领域语义网的应用提供便利。运用基础建模原语与OWL本体语言,构建包含语义网知识和语义的本体模型,读取与解析该本体模型后储存于数据库内。建立模糊贝叶斯网络模型,运用联结树信念传播算法作为其模糊概率推理算法,实现对所构建语义网本体模型内不确定性数据的推理,并对推理后的数据实施模糊化与去模糊化处理,获取最终高精度推理结果。结果表明,上述方法能够实现对各种情景模糊概率知识的推理,可行性较高,同时所提方法所融合的模糊化与去模糊化处理,可提升最终推理结果的精度,获得可信度与应用价值更高的推理结果,降低各领域语义网的实际应用难度。

Currently,some methods ignore analyzing ISA relations between ontologies,when calculating their semantic similarity,leading to low accuracy and low efficiency.Therefore,a method for calculating the semantic similarity between web ontologies based on ISA relation was proposed.Firstly,network link structure features,entity-tag dependence features and tag features were extracted.On this basis,a classification model was built to obtain ISA relations between web ontologies.Based on the ISA relationship,the information factor,distance factor,attribute factor,and hierarchy factor were linearly weighted.And then the semantic similarity was calculated.Finally,the overall effectiveness of the proposed method was verified by the test of calculation accuracy and calculation efficiency.