Summary: The sharing of data for mining has been proven beneficial in industry, but requires privacy safeguards. Some companies prefer to share their data for collaboration, while others decide to share only the patterns discovered from their data. The goal of these companies is to disclose only part of the knowledge and conceal a group of sensitive rules (competitive knowledge) that are paramount for strategic decisions. These rules must be protected before sharing and need to remain private. The challenge here is how to protect the sensitive rules without putting at risk the effectiveness of data mining per se. This work presents a framework for protecting sensitive knowledge in Association Rule Mining. The framework is composed of a set of heuristics and metrics to evaluate the effectiveness of these heuristics in terms of information loss and knowledge protection.

Publication year: 2006

Types of publication: Booklets

Unit: Embrapa Digital Agriculture