A data usage control system using dynamic taint tracking

Data analytics services are on the rise, fostered byan increasing number of wearables and industrial sensors whichare connected over the Internet. As a result, users who want totake advantage of these services are confronted with the challengeof keeping their private data and business secrets secure, whilestill providing the information required for the analytics serviceto operate. Traditional access and usage control do not solve thisproblem, as they only take binary access decisions, but do notenforce specific views on data sets. We propose a mechanism tocontrol the ways in which data may be processed, thereby limitingthe information which can be gained from data sets to the specificneeds of a service. The core of our approach is to model dataanalytics as a data flow problem and to apply dynamic taintanalysis for monitoring the processing of individual records.