Resumen de Methodology for building secure big data ecosystems
Data is becoming an increasingly important asset for organizations of all types and sizes. This importance derives from the information that can be extracted from it, information which allows daily activities to be performed and that improves decision-making by top management. That is why there is ongoing and continuous growth in the amount of data being managed. This data has different structures, and may need to be analyzed while it is being generated. To meet these needs, a new data analysis paradigm has emerged: Big Data. Every time a new technology emerges, it generates a set of opportunities that can be leveraged to enable advantage to be taken of it. It must also be acknowledged that each emerging technology may lead to the appearance of new security issues. Big Data ecosystems are no exception. In general, security issues related to Big Data are related to data privacy and confidentiality, and are due both to the complexity of these types of ecosystems and to the fact that they were not designed with security in mind. It is in an effort to address this problem that in this doctoral thesis we present a methodology for building secure Big Data ecosystems.
This methodology is composed of two main artifacts: on the one hand, a security reference architecture (SRA) for Big Data ecosystems, which abstracts the main components of this type of environment, while also integrating security concepts, and on the other hand, a general process that is based on the SRA and which aims to incorporate security in the main stages of the life cycle of a Big Data ecosystem by following the security-by-design approach. This process is composed of three sub-processes that address the analysis and design, implementation and operation stages. Our methodology is based on widely accepted industry standards and best practices, thereby improving its validity and applicability to any scenario. In addition, in this doctoral thesis we incorporated the creation of a set of security patterns specific to Big Data which can be applied during the process and that are based on the components defined in the SRA.
Finally, in order to check the feasibility and applicability of our proposal, we carried out a case study in a real Big Data scenario, a study which compares the approach of this already-implemented ecosystem with our methodology. This case study has allowed us not only to refine our proposal, but also to create a set of recommendations for the organization of the case study.
