Big Data Strategies for Agile Business

Bhuvan Unhelkar

• Editorial: Auerbach Publications
• Año: 2017
• ISBN:
  9781351646543
• Edición: 1
• Categoría: Business
• Categoría: Gestione dell'informazione

Agile is a set of values, principles, techniques, and frameworks for the adaptable, incremental, and efficient delivery of work. Big Data is a rapidly growing field that encompasses crucial aspects of data such as its volume, velocity, variety, and veracity. This book outlines a strategic approach to Big Data that will render a business Agile. It discusses the important competencies required to streamline and focus on the analytics and presents a roadmap for implementing such analytics in business.

ANALYTICS, PROCESSES, TECHNOLOGIES, ARCHITE CTURE, AND DATABASES WIT HIN THE BDFAB II

Chapter 3

Data Science—Analytics, Context, and Strategies

Chapter Objectives

Understanding data management, analytics, and strategies that are integral to developing data science within an organization

Dealing with different data types (audio, video, image or graphic, sensor, text and numbers, and mixed) and data characteristics (volume, velocity, variety, and veracity)

Establishing the life cycle and context (e.g., time and location) of a data point

Machine learning (ML) and the role of hex elements in establishing the context for a data point

Defining the concept of granularity of data and analytics (fine vs. coarse granular) and considering the factors influencing the optimum granularity level (OGL)

Discussing the various analytics categories and their role in enabling business agility

This chapter is mainly based on the second module of the Big Data Framework for Agile Business (BDFAB): data science—analytics, context, and technology. Specifically, as shown in Figure 2.1, this module draws attention to data, its various types and categories, and their utilization in analytics. Setting the context of a data point and the role of hex elementization in doing so is also discussed in this chapter. Furthermore, this chapter explains the crucial importance of granularity in data and analytics, and the setting of the OGL. Data science is presented as a discipline responsible for adopting and using Big Data in an iterative and incremental manner.

Data Science: Analytics, Context, and Strategies

Understanding the Importance of Data Science

Data science is a broad-ranging term that represents the technologies and analytics of Big Data. Additionally, though, data science can be understood as a discipline of utilizing technologies and analytics to convert data into actionable knowledge. Data science includes data mining, analytics (statistics), process modeling, machine learning (ML), parallel processing, and associated aspects of data management. The application of analytics to this data is the main step in arriving at insights. Therefore, data analytics remains at the core of data science. Data science, however, is closer to the business leadership and strategic decision making than data analytics. The evolution of data to actionable knowledge requires a specialist discipline that includes the study of data, its characteristics, its context in analytics, and eventually its value in business agility. These aforementioned activities require a wide coverage of various other disciplines within the organization and collaboration with many cross-functional teams. Therefore, the work of data science is interdisciplinary.

While analysis of data can focus on using the statistical expertise and management of data on the technical aspects, data science focuses on the strategic aspect of producing business value from data. This strategic aspect of data science requires domain knowledge of the industry where it is applied. For example, domain knowledge of the banking, finance, insurance, telecom, medical, and education industries is important in developing corresponding data strategies. Consider the following examples where data science provides value by combining the science of data with the domain knowledge:

Undertaking detailed analytics to suggest the pricing of an airline ticket with a degree of level of confidence in the prediction—requiring knowledge of the airline pricing strategies, as well as that of suitable data and analytics.

Analyzing a bank’s internal enterprise systems data and combining it with demographic metadata in order to identify potential loan defaulters based on knowledge of credit risks ¹ and associated regulations. This exercise cannot be conducted by someone with no knowledge of the banking and finance domain.

Predicting the risk of credit card fraud based on a wide range of micro (individual) and macro (group demographic) parameters and the personal credit market. This activity requires knowledge of financial fraud detection, in addition to data and its analysis.

Bringing together weather, soil, and economy data to predict the capacity for storing agricultural produce—requiring a combination of knowledge in multiple disciples of weather, agriculture, economy, and even logistics (for transportation).

Exploring the revenue trends and relating them to customer turnover for a hotel chain. Developing this data pattern requires knowledge of statistical analysis, as well as the hospitality domain.

Preparing production schedules and support logistics for the delivery of goods from a manufacturing organization requiring knowledge of production scheduling, together with the predictive aspect of data analysis and the associated supply chain.

Developing a promotion strategy in a democratic election process based on fine granular analytics on keywords and their relationship to voter demographics. Such promotion needs an understanding of the political and voting process as much as the analytical process.

Facilitating the organization of communities around common data and analytical interests (e.g., buying groups, political groups, and environmental groups). This requires an understanding of how communities are formed, what sustains them, the risks associated with their formation, and the value they provide.

Data science thus has wide-ranging applications in business decision making. Many internal organizational disciplines and functions provide input into data science. These are the disciplines of business strategies, project management, enterprise architecture, process modeling, solutions development, and quality assurance and testing. These disciplines complement those of data science. Figure 3.1 shows the effort involved in categorizing data, finding the correlation, undertaking analytics, and presenting the insights in an easy-to-use way for the end user.

Figure 3.1Data analytics, data categories (pools), and a subprocess for data transformation.

Data science considers the following in order to provide business value:

Key business outcomes desired by the business from adopting Big Data. This is a leadership activity that requires close collaboration with the overall senior leadership team of the organization.

Understanding the impact of Big Data–enabled processes on the complex dynamics of organizational structures and behavior. This requires an understanding of the way organizations are structured, how they change when the decision making is decentralized, and the way in which business processes can be modeled for optimization.

Current technology maturity of the organization, including its enterprise architecture. This point highlights the need to understand the current capabilities of an enterprise and how they can be enhanced to enable Big Data adoption.

Current business processes and how these processes will change (be reengineered), with the o...

Índice

• Cover
• Half-Title
• Title
• Copyright
• Dedication
• Series
• Contents
• List of Figures
• List of Tables
• Foreword
• Preface
• Acknowledgements
• About the Author
• Domain Terms and Acronyms
• SECTION I INTRODUCTION TO BIG DATA STRATEGIES AND OUTLINE OF BIG DATA FRAMEWORK FOR AGILE BUSINESS (BDFAB)
• SECTION II ANALYTICS, PROCESSES, TECHNOLOGIES, ARCHITE CTURE, AND DATABASES WIT HIN THE BDFAB
• SECTION III QUALITY, GRC, PEOPLE AND THEIR UPSKILLING, AND AGILE BUSINESS WIT HIN THE BDFAB
• SECTION IV CASE STUDIES IN BANKING, HEALTH, AND EDUCATION