| Summary: | 碩士 === 華夏科技大學 === 資訊管理系碩士在職專班 === 106 === In recent years, it has become common practice to publish some individual information to provide research and analysis such as for the purpose of scientific research, business marketing analysis, or public health policy promotion. Focusing on the analysis of these materials, it is possible to obtain valuable and useful information that can help the formation of the right decisions in all aspects. Usually these data can be used to represent in the graphical way, that is, the vertices in the graph represent real entities in the real world, and the edges in the graph represent the interaction between the entities pairs. However, it is unavoidable that individual information often contains personal privacy and sensitive information, but how to prevent a potential attacker from identifying the relationship between a particular vertex in the graph and individual entity of a real world to protect the individual privacy and sensitive information not to be exposed and infracted, which is a critical research topic for data providers. In existed anonymous technology, anonymities data either in the tabular or graphs-based, solve the problem of anonymization of data only from a single way.
In this paper, we propose a semantic-based anonymization algorithm for data, and that uses the ontology of the individual entity, we also considers the characteristics of the tabular and the bipartite graphic data, carries on the operation of the anonymization of the graphic data. It can efficiently carry out secure and strict (k, l)-grouping on data. After the data was anonymized, that can carry out complex query and answer capabilities.
|
|---|
