BIG DATA ANALYTICS TYPES

The following four types of big data analytics were distinguished [9] (Figure 1.3):

• Descriptive Analytics: It consists of asking the question: “What is happening?’' It is a preliminary stage of data processing that creates a set of historical data. Data mining (DM) methods organize data and help uncover patterns that offer insights.

• • Diagnostic Analytics: It consists of asking the question: “Why did it happen?” Diagnostic analytics look for the root cause of a problem. It is used to determine why something has happened. This type attempts to find and understand the causes of events and behaviors.
• • Predictive Analytics: It consists of asking the question: “What is likely to happen?” It uses past data in order to predict the future. It is all about forecasting. Predictive analytics uses many techniques such as DM and artificial intelligence to analyze the current data and make scenarios of what might happen.
• • Prescriptive Analytics: It consists of asking the question: “What should be done?” It is dedicated to finding the right action to be taken. With descriptive analytics providing historical data and predictive analytics, helping forecast what might happen, prescriptive analytics use these parameters to find the best solution.

FIGURE 1.3 The 3Ps that describe big data purpose.

If the people now had not been living in an era where they produce a lot of data, the important question to ask would have been, “Will they adopt a big data approach?” The answer can be summarized in three main reasons: First, the exponential increase in the number of connected users, connected smartphones, connected tablets, connected glasses, and as a result, connected objects. In addition, the individuals have become more reliant on terms of quality and costs. Finally, if so much data is being produced, data can be stored in different storages, especially with the digitization of society.

Certainly, big data plays a very important role for governmental organizations, private and multinational companies, whatever their field of activity, it applies to all types of companies, large or small, but with a necessary condition: it has to generate large volumes of data.

At first, big data was used by a specific sample of companies such as banks for credit card transactions and financial market-related uses, by telephone companies for telephone call records, and by e-commerce sites (e.g., Amazon and eBay) to improve online services. Although big data started in specific industries, it is now available to everyone, even small SMEs [10].

The value chain, the concept introduced by Porter [16], refers to a set of activities carried out to create added value at each stage of product design or to provide a service to its customers. Similarly, the data value chain refers to the framework that deals with a set of activities aimed at creating value from available data. It can be divided into four essential phases: data integration, data storage, data manipulation, data security, data analysis, and decisionmaking [11].

1.4.1 BIG DATA CLASSIFICATIONS

Big data can be classified into the following three categories [12]:

• • Structured Data: It refers to any kind of data which is stored in relational databases and spreadsheets that reside in a fixed field within a record or file.
• • Unstructured Data: The phrase unstnicmred data usually refers to information that doesn’t reside in a traditional row-column database. As you might expect, it’s the opposite of structured data-the data stored in fields in a database.
• • Semi-Structured Data: It is data that hasn’t been organized into a specialized repository, such as database, but even so, has associated information such as metadata, which makes it more amenable for processing than raw data.

The success of a big data project is largely linked by its architecture and its conect infrastructure, so the big data architecture is based on four components, as mention in Figure 1.4.

To summarize, an integration which consists of loading the volume of data onto storage media and then storing them in order to manipulate them, including the processing objective and better extract a reliable and conect result [13].

FIGURE 1.4 Big data architecture.