Ripple-Effect Analysis of Ontology Evolution Based on CK-Modes Algorithm
Jianghe Gong, Haidong Fu · 2020
Ontology users can accurately and comprehensively analyze the construction and evolution of the ontology by quantitatively analyzing the Ripple-Effect of the concepts or entities in the ontology, and provide a basis for the subsequent updating of the ontology. Existing methods quantify the Ripple-Effect in ontology, but do not consider the impact of ontology semantic structure on the Ripple-Effect. Existing methods preliminary quantify the Ripple-Effect in ontology, but do not consider the impact of ontology semantic structure on the Ripple-Effect. Aiming at the above problems, this paper proposes a method for analyzing the Ripple-Effect of ontology based on CK-Modes clustering. This method transforms the ontology into an ontology labeled attribute graph, cut the ontology according to the CK-Modes clustering algorithm, multiplied by different quantization coefficients according to different cutting modules, thus quantify the semantic from the perspective of semantic structure. Then, the transformation from the ontology to the semantic relation matrix is completed through the quantified semantic relations. Finally, the effect of each concept or entity is calculated by the improved Freud algorithm. The experimental result show, CK-Modes clustering-based ontology evolution Ripple-Effect analysis method can more accurately measure the concept or entity Ripple-Effect during the ontology evolution process. At the same time, the results of the measurement of the common nodes in different versions are more stable.