Abstract: This paper analyzes the meaning of data mining in shipwreck data, improves the Apriori algorithm and applies it into finding frequent patterns of shipwreck data warehouse which is organized as a snowflake schema while improving direct rule graph to visualize the association rules. Research result shows that data mining technology for further research on historical shipwreck data can overcome the limitations of the traditional statistics and analysis and can mine a lot of knowledge so that some references can be provided for future navigation safety.

Key words: Apriori, Association rule, Data mining, Shipwreck, Visualization

摘要： 分析了建立海难数据仓库的意义，提出了海难数据仓库的雪花模型，对Aprioir算法进行了改进，用改进后的算法实现了海难数据的关联规则和频繁模式挖掘，用改进的有向图方法实现了关联规则的可视化表示。结果表明，利用数据挖掘技术对海难历史数据作深层次分析，克服了传统统计分析方法的局限性，可挖掘出大量的知识，为以后的航海安全提供借鉴。

关键词: Apriori, 关联规则, 数据挖掘, 海难, 可视化

CLC Number: 

• TP311