海上救助打捞知识图谱构建及应用

Construction and Application of Marine Rescue and Salvage Knowledge Graph

针对海上救助打捞(MRS)领域知识体系零散及海量救助打捞案例数据无法有效挖掘和利用的问题,通过对救助打捞事故案例和标准规范文件的详细分析,获取5类事故对象、19个场景及每个场景所需的关键救助打捞能力。基于这些场景和能力需求之间的关系,采用自顶向下与自底向上方式相结合构建MRS内部实体和属性图,对数据进行结构化处理并映射到设计图谱,并结合决策推荐规则实现知识图谱的构建和应用。研究结果表明:MRS知识图谱能够有效进行救助打捞知识的检索查询、知识展示与信息统计可视化及辅助决策方案推荐,为相关领域的作业提供更加系统化和智能化的支持工具。

Aiming at the problem that the knowledge system in the field of Marine Rescue and Salvage(MRS)is scattered and the massive case data cannot be effectively mined and utilized,through the detailed analysis on rescue and salvage accident cases and standard specification documents,five types of accident objects,nineteen scenarios,and key rescue and salvage capabilities required for each scenario are obtained.Based on the relationship between these scenarios and capacity requirements,the top-down and bottom-up combination is used to construct the internal entity and attribute graph of MRS.The data is structured and mapped to the design graph,and the construction and application of the knowledge graph is realized by combining the decision recommendation rules.The research results show that the MRS knowledge graph can effectively carry out the retrieval and query of the rescue and salvage knowledge,the visualization of knowledge display and information statistics,and the recommendation of auxiliary decision-making schemes,which provides a more systematic and intelligent support tool for operations in related fields.