本体映射作为实现多本体间相互操作的重要手段,已广泛应用于诸多领域。利于k-部排序学习算法得到最优排序函数,从而将多本体结构图中每个顶点映射成一个实数,通过比较来自不同本体顶点对应实数间的差值判断概念间的相似程度。实验表明:该方法对于在不同本体间建立映射是有效的。
Ontology mapping, as an important means of multi-ontology mutual operations, has been widely used in many fields. The optimized ranking function was obtained by using k-partite ranking learning method. The function mapped the vertices of the graph into real numbers. The similarities of the concepts from different ontologies were calculated by comparing the difference of their corresponding values. The experiment shows that the method for building mapping among different ontologies is efficient.