Data Privacy Protection for the Cloud

Abstract

Privacy is for a long time a concern when data is being discussed. Nowadays, with an increasing amount of personal and confidential data being transmitted and stored online, data curators have to assure certain guarantees of data protection and privacy. This Master Dissertation presents a background of anonymization and concealing techniques. Their characteristics and capabilities are described, as well as tools to implement and evaluate anonymization and concealing. The evaluation of the applicability of the DNA-inspired concealing algorithm is the main objective of this work. Usually, various metrics are used to measure aspects like risk or utility of the anonymized data. This work presents a new approach of evaluating how well concealed is the data. By using the Cosine Similarity as a measure of similarity between the private and concealed data, this metric proves its worthiness not only in information retrieval or text mining applications but also in the analysis of concealed or anonymized files. Nowadays there is a continuously growing demand for Cloud services and storage. The evaluation in the Master Dissertation is directed to find how suitable is the application of the DNA-inspired concealing algorithm over the data being stored or transmitted in the Cloud. The evaluation is made by analyzing the concealing results as well as the performance of the algorithm itself. The application of the algorithm is made over various texts and audio files with different characteristics, like size or contents. However, both file types are unstructured data. Which is an advantage for being accepted as an input by the algorithm. Unlike many anonymization algorithms which demand structured data. With the final results and analysis, it will be possible to determine the applicability and performance of the referred algorithm for a possible integration with the Cloud.