Prevention and Detection of Sensitive Information Exposure by Advanced Context Based Data Mining Techniques

In today's world, data security has become a major issue of concern to be addressed. Information Exposure or Leakage is one of the prominent way through which data security is compromised. The data exposure may occur accidentally or intentionally. So a model is required to prevent such kind of scenario. Existing methodologies work on specific phrases or are statistical based techniques. -based methods are not accurate as they do not take context into consideration, and statistical methods ignore the analysis of the text. Thus a new context based approach is introduced to prevent data leakage. This approach anchors the advantages of both the statistical approach as well as based approach. This approach is basically divided into two phases-Drilling stage and Identification stage. The first one deals with the clustering of documents and the latter phase determines the confidentiality. As this approach offers the best of both worlds, it appears that this model is the superior one.