Hiding Sequential and Spatiotemporal Patterns

The process of discovering relevant patterns holding in a database was first indicated as a threat to database security by O'Leary in. Since then, many different approaches for knowledge hiding have emerged over the years, mainly in the context of association rules and frequent item sets mining. Following many real-world data and application demands, in this paper, we shift the problem of knowledge hiding to contexts where both the data and the extracted knowledge have a sequential structure. We define the problem of hiding sequential patterns and show its NP-hardness. Thus, we devise heuristics and a polynomial sanitization algorithm. Starting from this framework, we specialize it to the more complex case of spatiotemporal patterns extracted from moving objects databases. Finally, we discuss a possible kind of attack to our model, which exploits the knowledge of the underlying road network, and enhance our model to protect from this kind of attack. An exhaustive experiential analysis on real-world data sets shows the effectiveness of our proposal.

1 Introduction

Knowledge hiding refers to the activity of concealing sensitive patterns holding in a database that is going to be published. If the data are published as it is, the sensitive patterns may be surfaced by means of data mining techniques. Knowledge hiding is usually obtained by sanitizing the database in such a way that the sensitive knowledge can no longer be inferred, while the original database is changed as little as possible. Different approaches for knowledge hiding have emerged over the years, mainly in the context of association rules mining. But nowadays, the real-world applications call for more advanced analysis of more structured kind of data.