Abstract:

From knowledge discovering of information, this paper proposes a water eutrophication risk analysis modeling method based on Probabilistic Relational Model(PRM). This modeling method uses storage structure and content of multi-relational data to learn and mine, then builds a network topology model based on PRM. Example analysis results show that PRM model can easily explain and analyze the correlations among all effects, the modeling method can discover the talent risk of water eutrophication by analyzing the historic data, to some extent it can support decision-making for water environment management and water pollution control.

Key words: multi-relational data mining, Probabilistic Relational Model(PRM), eutrophication risk analysis

摘要：

从信息的知识发现角度出发，提出基于概率关系模型(PRM)的水体富营养化风险分析建模方法。该建模方法利用多关系数据的存储结构和存储内容对数据进行学习与挖掘，构建具有网络拓扑结构的PRM模型。示例分析结果表明，PRM模型易于解释与分析水体中各种影响因素间的相关性，该建模方法可通过分析历史数据发现水体富营养化的潜在风险，为库区水环境管理与水污染防治提供科学依据。

关键词: 多关系数据挖掘, 概率关系模型, 富营养化风险分析

CLC Number: 

• TP181