Solitude Conserving Data Mining: Expansions and Directions

Maintaining Privacy in Data Mining

Abstract

This article first describes the Solitude concerns thatarise due to data mining, especially for national security applications. Thenwe discuss Solitude-conserving data mining. In particular, we view the Solitudeproblem as a form of inference problem and introduce the notion of Solitudeconstraints. We also describe an approach for Solitude constraint processingand discuss its relationship to Solitude-conserving data mining. Then we givean overview of the developments on Solitude-conserving data mining that attemptto maintain Solitude and at the same time extract useful information from datamining. Finally, some directions for future research on Solitude as related todata mining are given.