A Framework for Inference-Directed Data Mining

Abstract

This paper presents a second-path inference-detection approach based on association cardinalities.* It is applicable to the detection of second paths that do not involve functional dependencies or foreign keys. It provides for an analysis sieve that begins with the analysis of an object model of the database. The goal of the analysis is to detect cases in the database in which a small number of values in the target entity can be associated with a single value in the anchor entity. The number of values is called the association cardinality from anchor to target. Inference vulnerabilities occur for cases of small association cardinalities. The analysis sieve processes the data model of the database to detect cases of small association cardinality. For cases with high cardinality associations, the sieve mines the database to detect cases of small instance-level association cardinalities.

Protecting Privacy in Data Release

“…In recent years, many other proposals have been presented to deal with inference channels, aiming to a general and strong formulation of the problem in order to find formal and automated models and frameworks to provide protection [18,34,36,47,64,77]. Hinke and Delugach [47,[63][64][65] propose a solution for an automated analysis of inferences in general purpose databases, addressing the representation of external knowledge as well.…”

“…Hinke and Delugach [47,[63][64][65] propose a solution for an automated analysis of inferences in general purpose databases, addressing the representation of external knowledge as well. The proposed method is based on a graph representation to locate inference channels, representing the knowledge needed for the problem, such as data items, relationships between them, domain knowledge and data sensitivity.…”

Bayesian Methods Applied to the Database Inference Problem

“…The inference problem fits into the bigger scheme of data mining [18], [15], [12]. With this in mind, we use the data mining (learning) technique of Bayesian networks to analyze the inference problem, meeting conditions one and two as above.…”

The Security Problem Against Inference Attacks on Object-Oriented Databases

“…In [15], a quantitative measure of inference risk is formally defined. Imprecise inference with external, common sense knowledge can be regarded as data mining [7,11]. [14] focuses on both precise and imprecise inference in OODBs.…”