The term “big data” synthesizes the problems and technologies that arise with the spread of big volumes of data. Usually the five Vs model is used to characterize big data problems: velocity (speed of data in and out), volume (scale of data), variety (heterogeneity in data types and data sources), veracity (uncertainty of data), and value (reasonable cost). However, these are just common characteristics for a large range of problems.
This book offers an introduction to several big data problems and situations. It also covers the technical solutions used in such situations. The book deals with a varied set of issues. Technical frameworks such as Hadoop and architectures such as MapReduce are studied. Security, in several of its manifestations, receives a lot of attention, which is logical, as security is itself a main concern nowadays. The book presents big data applications in different areas, such as finance, the environment, and science.
Chapter 1 deals with problems that appear when composing event streams. Chapter 2 is an extensive overview of big data applications, with special attention to the Hadoop ecosystem, one of the technologies that comes to mind when talking about big data. It also offers an introduction to the concepts needed to understand it. Chapter 3 presents the effects of noisy labels in classification processes associated with network traffic. In chapter 4, the security concerns that appear with cloud computing, such as confidentiality, integrity, and data privacy, are treated. Data leaks are the target of chapter 5. Despite this also being a security concern, this chapter deals with data leaks during data flows and the techniques used to discover them in network channels. Security comes again to the arena in chapter 6, this time in the form of storage security concerns: attacks on stored data, and how to detect and prevent them. Attacks are also covered in chapter 7. However, in this chapter, the attacks studied are those that damage computation time by compromising nodes in Hadoop clusters. In chapter 8, mechanisms for data encryption, privacy preservation, and trust management are presented.
In chapter 9, the perspective changes: it is no longer about problems that can be found with big data, but contexts in which big data can be present, for example, clinical decision support. In chapter 10, the technologies used for geospatial datasets are presented, pointing out the big data properties present in this area. Chapter 11 discusses the applications of big data in the financial sector. Finally, technical solutions for business analysis can be found in chapter 12.
Every chapter follows a similar structure: the problem dealt with is presented first, emphasizing the properties that characterize it as a big data problem; an overview of the available solutions follows; and a small conclusion ends the chapter. This is valuable in a collective book such as this one, with different authors for every chapter. Perhaps chapter 2 is a bit of an exception to this rule, as it is an extensive overview of big data applications. Nevertheless, as it can be used as a support for the rest of the chapters, it does not interfere with the coherence of the book (but why is it not the first chapter?).
The book is a nice tool to get in touch with the big data problems that can appear in different areas and obtain an introduction to the main solutions used to cope with them (algorithms, techniques, architectures, tools, and so on). Each chapter offers an overview for every issue’s problems and solutions. Of course, the fine-grained details of each solution should be searched elsewhere. On the other hand, to make the most of this book, a good background in computing is advisable. This is a technical book; without the necessary background, I do not think that the essence of the problems and solutions presented could be understood. Thus, the target audience is computer experts.
