一种可降低结构性变化的本体演化算法  

An Ontology Evolution Algorithm for Reducing Structural Changes

作　　者：陆晨阳 袁嵩[1,2] 高峰 顾进广 LU Chen-yang;YUAN Song;GAO Feng;GU Jin-guang(School of Computer Science and Technology,Wuhan University of Science and Technology,Wuhan 430065,China;Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial,Wuhan 430065,China;WISDRI CCTEC Engineering Co.,Ltd.,Wuhan 430223,China)

机构地区：[1]武汉科技大学计算机科学与技术学院,湖北武汉430065 [2]湖北省智能信息处理与实时工业系统重点实验室,湖北武汉430065 [3]中冶南方连铸技术工程有限责任公司,湖北武汉430223

出　　处：《计算机技术与发展》2024年第6期45-52,共8页Computer Technology and Development

基　　金：国家自然科学基金(U1836118);国家重点研发计划(2022YFC3300800);武汉市重点研发计划(2022012202015070);武汉东湖新技术开发区“揭榜挂帅”项目(2022KJB126);湖北省教育厅科学研究计划项目(B2022016);武汉市知识创新专项-曙光计划项目(2023010201020409)。

摘　　要：在Web语义环境中,如何降低本体在不断演化的过程中的结构性变化是一项具有挑战性的任务。针对这一问题,提出了一种本体演化算法,旨在降低本体演化过程中的结构性变化,从而提高本体的稳定性和可维护性。与传统的本体演化方法主要关注实体间关系不同,该文在计算本体演化代价时在实体频率和实体间关系频率的基础上综合考虑了实体之间的概念关系,从而实现了对实体影响的更细粒度分析。此外,还提出新的附加演化策略,以便在演化过程中选取最优的演化策略,保留更多复杂的语义关系以实现最小结构性变化。实验结果表明,提出的演化方法在处理涉及复杂语义关系的本体演化时表现出色,有效地降低了演化代价并提高了本体演化前后的结构相似度。这些研究成果在Web语义环境中对本体进行演化和更新方面具有重要意义,为本体工程领域提供了有益的改进方向和实践经验。In the Web semantic environment,how to reduce the structural changes of ontologies during their continuous evolution is challenging.Aiming at the problem,we propose an ontology arithmetic algorithm to reduce the structural changes during ontology evolution and improve the stability and maintainability of ontologies.Unlike traditional ontology evolution methods that mainly focus on inter-entity relationships,we integrate the conceptual relationships between entities based on entity frequencies and inter-entity relationship frequencies when calculating the ontology evolution cost,thus realizing a more fine-grained analysis of entity impact.In addition,we propose additional evolution strategies to select the optimal evolution strategy during the evolution process and retain more complex semantic relations to achieve minimal structural changes.Experimental results show that the evolutionary approach proposed performs well in dealing with ontology evolution involving complex semantic relations,effectively reducing the evolutionary cost and improving the structural similarity before and after ontology evolution.These research results are of great significance in evolving and updating ontologies in Web semantic environments and provide useful improvement directions and practical experience in the field of ontology engineering.

关 键 词：本体 本体演化 语义关系 演化策略 结构性变化 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]