•  
  •  
 

Journal of System Simulation

Abstract

Abstract: The construction of the unified expression model of battlefield situational information is challenging due to the complexity of data sources and the significant differences in data structures and expression methods. Ontologies, as semantic conceptual models, are often used to describe concepts, relationships, and attributes within knowledge domains. An ontology construction method for the battlefield situational information domain based on a top-down and bottom-top integration is proposed. The top-down method is used to construct the upper ontology, in which a conceptual hierarchy model with a clear top-down structure is designed to establish the hierarchical relationships and semantic associations. A bottom-up method is used to construct the lower ontology, and a novel double aggregation algorithm is proposed for its extraction. The feature representation of entities is obtained by using TF-IDF. The ontology is initially abstracted by using K-means++ algorithm, and an agglomerative hierarchical clustering algorithm is applied to further refine the ontology concepts. The ontology model BattleOnto is constructed by using OWL and Protégé, which can support the decision making.

First Page

2469

Last Page

2487

CLC

TP391

Recommended Citation

Zhou Cong, Zhou Sihang, Huang Jian, et al. A Method for Battlefield Situation Information Ontology Construction Based on Top-down and Bottom-up Integration[J]. Journal of System Simulation, 2024, 36(10): 2469-2487.

Corresponding Author

Huang Jian

DOI

10.16182/j.issn1004731x.joss.23-0665

Share

COinS