基于改进遗传规划算法的本体匹配方法

打开文本图片集

中图分类号：TP301.6 文献标志码：A 文章编号：2097-3853（2025）04-0344-10

Ontology matching method based on improved genetic programming algorithm

SHEN Jianchao¹，DAI Ketao²，XUE Xingsi³，LI He²，LYU Qing² （1. Xiadian Coal Mine of Lu'an Chemical Group Cilinshan Coal Industry Co.，Ltd.，Changzhi O46299，China; 2. School of Electrical and Power Engineering，Taiyuan University of Technology，Taiyuan O3OO24， China; 3.School of Computer Science and Mathematics，Fujian University of Technology，Fuzhou 350118，China）

Abstract：Inspired by the successul application of genetic programming （GP）in the field of feature construction，an ontology matching method based on improved genetic programming algorithm （ESMPD-GP）is proposed.In order to solve the problem of GP individual simplification pressure and avoid the algorithm falling into local optimum in the evolution process，ESMPD-GP adopts two new algorithm components，namely，dualobjective elite selection mechanism and population diversity enhancement strategy.The former selects the individual set with thestrongestdiversity in the population by considering both individual fitness and edit distance.The later introduces excellent solutions to simplify the pressure by creating additional populations， increasing population diversity and preventing premature convergence.Experimental results based on Benchmark and Anatomy test sets on the Ontology Alignment Evaluation Initiative（OAEI）show that the proposed method outperforms existing state-of-the-art methods with an average f-measure of O.96 on Benchmark.

Keywords：ontology matching；genetic programming； elite selection；population diversity enhancement

计算机领域的本体（ontology）作为语义网（semanticweb，WB）基础技术，通过类、属性、实例等元素对领域知识进行形式化表达[1]，支撑自然语言处理、智能搜索及机器学习等应用。（剩余13353字）