Anchor-based ontology partitioning and Genetic Programming with Relevance Reasoning for large-scale biomedical ontology matching

Created by W.Langdon from gp-bibliography.bib Revision:1.8414

@Article{Sun:2025:eswa,

• author = "Donglei Sun and Qing Lv and Pei-Wei Tsai and Xingsi Xue and Kai Zhang",
• title = "Anchor-based ontology partitioning and Genetic Programming with Relevance Reasoning for large-scale biomedical ontology matching",
• journal = "Expert Systems with Applications",
• year = "2025",
• volume = "270",
• pages = "126445",
• keywords = "genetic algorithms, genetic programming, Large-scale biomedical ontology matching, Anchor-based hierarchical partitioning, Relevance Reasoning, Task termination prediction",
• ISSN = "0957-4174",
• URL = "https://www.sciencedirect.com/science/article/pii/S0957417425000673",
• DOI = "doi:10.1016/j.eswa.2025.126445",
• abstract = "In the field of biomedicine, the heterogeneity of features in biomedical ontologies (BO), along with their complex interrelationships and large scale, poses significant challenges for entity matching. This paper presents a novel Genetic Programming (GP)-based method for parallel matching of BO, aimed at addressing the efficiency and quality of alignments associated with large-scale biomedical ontology matching (BOM). By introducing the Compact Relevance Reasoning GP-based algorithm (CRRGP), this study enhances entity alignment by leveraging the implicit structural relationships inherent among entities in partitioned sub-ontologies, thereby significantly improving the efficiency and quality of matching tasks. Furthermore, an innovative anchor-based hierarchical partitioning algorithm is introduced to tackle the spatiotemporal complexities associated with large-scale BOM. This algorithm effectively retains the features of partitioned entities while promoting collaborative reasoning among sub-ontologies. To optimise computational resources and ensure timely task completion, we developed a novel evaluation metric for predicting task termination. Evaluations conducted using the BO test cases from the Ontology Alignment Evaluation Initiative (OAEI) demonstrate that this approach not only reduces redundant computations but also ensures reliable matching outcomes. This research advances methodologies in BOM and contributes to more effective data interoperability within the biomedicine domain",

}

Genetic Programming entries for Donglei Sun Qing Lv Pei-Wei Tsai Xingsi Xue Kai Zhang

Citations