Identification and Evaluation of Concepts for Privacy-enhancing Big Data Analytics Using De-Identification Methods on Wrist-Worn Wearable Data

Abstract

Nowadays, Wearables like Smartwatches do not only serve as a tool for sports activity tracking, but they also perform several health-related measurements on the individuals. Different sensors collect information about the user's heart rate, blood oxygen saturation (SpO2), blood pressure, body temperature and even clinically tested ECGs can be conducted on recent devices. The service providers use this data to detect health issues like cardiac dysrhythmia, atrial fibrillation and sleep apnea. Additionally, analytical functions are performed on the aggregated datasets in order to achieve general product improvements, support transport infrastructure planning and to determine correlations within the data to give the users improved training and health recommendations.

In contrast to that, health-related data is among the most sensitive data a human being can reveal. Therefore, it comes to an important trade-off between the users' data privacy and the data utility which enables analytical operations.

To perform Big Data Analytics, service providers rely on cloud infrastructure, which are a central point of attack for adversaries. A data leak could lead to the disclosure of personal health details which is a major privacy concern. 

To ensure data privacy in cloud environments, a solution based on a privacy gateway is envisioned. The gateway will use de-identification techniques to ensure that sensitive data is not stored in plain text in the cloud while at the same time enabling as many analytics functions as possible.

The goal of this thesis is to investigate the State of the Art of De-Identification methods for Privacy-enhancing Big Data Analytics. Additionally, requirements and trends for analytics of wrist-worn Wearable Data in the Cloud are examined. These insights are then used to develop a concept which enables Data Privacy for wrist-worn Wearable Data in the Cloud based on De-Identification methods.

Keywords: 

Big Data Analytics, Data Privacy, De-Identification, Wrist-Worn Wearables, Data-centric Security, Health Data

Research Questions

RQ1: What is the state of the art of approaches using de-identification methods for privacy-enhancing Big Data Analytics and how can they be distinguished from other approaches?

RQ2: What are requirements for privacy-enhancing analytics of wrist-worn wearable data in the cloud?

RQ3: What are concepts enabling data privacy for wrist-worn wearable data in the cloud based on de-identification Methods?

References