Privacy Preservation in Information System

Information System

The basic objective of inductive learning and data mining is to learn the knowledge for classification. However, in real world problems, we may not be faced with simply classification. One such problem is the ordering of objects. On the contrary, we are interested in hiding sensitive associations that is present in an information system. Before we discuss, various privacy preservation techniques to hide sensitive associations, one must know about an information system. An information system contains a finite set of objects typically represented by their values on a finite set of attributes. Such information system may be conveniently described in a tabular form in which each row represents an object whereas each column represents an attribute. Each cell of the information system contains an attribute value. Now, we define formally an information system as below.

An information system is defined as a quadruple where is a finite nonempty set of objects called the universe, is a finite nonempty set of attributes, is a nonempty set of values for , is an information function. For example, consider a sample information system as presented in Table 1 in which represents a nonempty finite set of objects; and A = {Humidity, Windy, Temperature} be a finite set of attributes. The information system presented in Table 1 is a qualitative system, where all the attribute values are discrete and categorical (qualitative).