基于本体构建技术的食品安全事件数据清洗研究

高质量的数据不仅为后续知识挖掘提供了最有效的数据源而且保证了所挖掘知识的可信度和可用度。为了确保基于食品安全事件数据所挖掘知识的有效性和真实性，本文提出了一种基于本体构建技术的食品安全事件数据清洗流程，并对所清洗的数据进行了相应的评价。本文所清洗的数据是酒鬼酒塑化剂这一食品安全事件数据，在具体清洗过程中，基于斯坦福大学医学院开发的七步法，通过Protégé这一工具，构建了相应的食品安全事件清洗本体。该研究为探究领域化数据的清洗工作提供了一种可行性的操作方案，并且对于所清洗的数据增加了一定的语义知识。

Research on the Food Safety Event Data Cleaning Based on Ontology Construction Technology

The high quality data not only provides the most effective data sources but also ensures the credibility and availability of the mining knowledge for subsequent knowledge mining. In order to ensure the effectiveness and authenticity of the knowledge of data mining based on food safety events, this paper puts forward a kind of food safety event data cleaning process based on ontology construction technology, and evaluates the data cleaning. The data used for cleaning is produced by the food safety event of Jiuguijiu plasticizer. In the specific cleaning process, the cleaning ontology of food safety event by Protégé is constructed based on the seven-step method developed by the School of Medicine of Stanford University. A feasible operation scheme for exploring the cleaning of domain-specific data is provided, and some semantic knowledge to the cleaned data is added.