Resenha dos capítulos 1 e 8 do livro Business Intelligence e Análise de Dados

C H A P T E R 1 ?

An Overview of Business Intelligence, Analytics, and Data Science ?

The business environment is undergoing constant and increasing changes, demanding organizations to be agile and make quick, innovative decisions in response to complex conditions. These decisions often rely on substantial data, information, and knowledge, necessitating rapid processing, frequently in real time, with computerized support. This book focuses on business analytics as computerized support for managerial decision-making, emphasizing theoretical foundations, commercial tools, and techniques. It introduces analytics techniques through an EEE approach (Exposure, Experience, and Exploration), providing exposure to inspire students to learn from real-world applications. The book presents fundamental techniques, details their construction and use, and offers examples of various software tools. Not limited to any specific tool, it encourages readers to explore analytics potential in their domains, providing exercises directing them to resources like Teradata University Network for team-oriented exercises.?

?

Learning analytics is a crucial skill addressed in this book, particularly in its application to business problem-solving. The focus extends to sports analytics, the art and science of leveraging data for sports-related decision-making. This includes decisions on player recruitment, compensation, team composition, training, health management, and strategic choices during games. Sports analytics also influences business aspects like ticket pricing and competitor analysis, emphasizing its growing importance within the broader analytics field. Considering the significant revenue sports generate—about $145 billion annually, with additional sums from legal and illegal gambling—analytics spending remains relatively modest at $125 million in 2014. Nevertheless, it is expected to grow substantially, reaching $4.7 billion by 2021.?

?

The popularization of sports analytics, notably through Michael Lewis's 2003 book "Moneyball" and the subsequent 2011 film featuring Brad Pitt, highlighted the transformative impact of data on sports management. The narrative focused on Oakland A's general manager Billy Beane, who strategically used analytics to revitalize a struggling team. This approach involved prioritizing unconventional player metrics, leading to successful draft choices and playoff appearances in 2002 and 2003.?

Sports analytics now pervades various aspects, categorized into front-office and back-office applications. Front-office analytics encompass fan behavior analysis, predictive models for ticket sales, and sentiment analysis of fan tweets—a parallel to traditional customer relationship management (CRM). Financial analysis, addressing factors like salary caps, is also vital.?

Back-office applications extend to individual athletes and team dynamics. For players, analytics contribute to recruitment, scouting, strength, fitness, development, and injury prevention. Team analytics encompass strategic decisions, competitive assessments, and optimal roster choices for diverse on-field situations.?

Three sports organizations serve as illustrative examples of effective data-driven analytics enhancing sports operations, mirroring how analytics have transformed decision-making in traditional industries.?

Beyond business analysts, coaches, trainers, and performance experts, various sports professionals, from golf groundskeepers monitoring PGA tournament conditions to referees in baseball and basketball tracking their calls, rely on data. Virtually every facet of sports is influenced by increased data availability, particularly from sensors.?

The business analytics skills acquired from this book are applicable to sports contexts. For those wishing to delve deeper, the Sports Analytics section of the Teradata University Network (TUN) offers valuable resources. This free platform provides insights into sports analytics, recommended readings, compilations of publicly available datasets for analysis, student projects, and interviews with sports professionals leveraging data and analytics in their roles. Wishing you success in your analytics learning journey!?

The opening vignette exemplifies how industries utilize analytics to generate reports, predict future scenarios, and make informed decisions. This necessitates extensive data collection and analysis, evolving from traditional payroll functions to advanced managerial applications. Information technology, deemed crucial by executives, has transitioned from transaction processing to cloud-based analytics tools like data warehousing, data mining, and dashboards, facilitating modern management.?

Today's managers rely on high-speed, networked information systems, often accessible via mobile devices, for decision-making. Automation has streamlined routine decisions, diminishing the need for managerial intervention. Developments, including group communication tools and improved data management, have further propelled decision support and analytics. Collaboration across locations is facilitated, particularly crucial in the supply chain, enhancing responsiveness to market changes. Advanced data management enables quick and secure storage, search, and transmission of diverse data types, fostering efficient decision-making processes.?

?

Managing extensive data warehouses and Big Data involves employing special methods such as parallel computing, Hadoop/Spark, etc., to organize and mine vast datasets. The costs associated with data storage and mining are rapidly decreasing. Big Data technologies allow organizations to analyze massive, diverse data from various sources, offering unprecedented insights into organizational performance.?

Analytical support, facilitated by increased data and advanced technologies, enhances the evaluation of alternatives, improves forecasts, and enables quick risk analysis. Decision makers can collect expert opinions from remote locations, perform complex simulations, assess scenarios, and evaluate impacts economically.?

Overcoming cognitive limits in processing and storing information is addressed through computerized systems that quickly access and process extensive information, surpassing human cognitive capabilities. Knowledge management systems, utilizing technologies like text analytics and IBM Watson, harness vast stores of organizational information, providing formal and informal decision support to managers.?

Anywhere, anytime support is a transformative change, allowing managers to access, analyze, and interpret information using wireless technology from any location. This shift, fueled by mobile technologies, social media platforms, and analytical tools, has elevated information systems (IS) support for managers since the late 1960s, intensifying since the mid-1990s. This growth extends to consumers, emphasizing data-driven decision support for a broader audience. The subsequent study covers an overview of BI technologies and expands into various types of analytics.?

In the 1970s, information systems primarily provided structured periodic reports for decision-making, leading to the development of Management Information Systems (MIS) to address changing business needs. Decision Support Systems (DSS) emerged, aiming to solve unstructured problems through interactive computer-based systems. Operations Research (OR) models, a precursor to prescriptive analytics, addressed complex optimization problems.?

The late 1970s and early 1980s saw the rise of rule-based expert systems, capturing expert knowledge for intelligent decision support. The 1980s introduced Enterprise Resource Planning (ERP) systems, integrating transactional data across organizational units into consistent relational database management (RDBM) systems. This integration improved data capture, storage, and consistency, fostering the emergence of on-demand business reporting.?

The 1990s witnessed the development of Executive Information Systems (EIS) with graphical dashboards and scorecards for executives. Data Warehouses (DW) became a central repository, supporting business reporting and decision-making. In the 2000s, DW-driven Decision Support Systems (DSS) evolved into Business Intelligence (BI) systems, addressing the need for current, digestible information. Real-time and right-time data warehousing emerged to tackle information latency.?

With growing data volumes, data mining and text mining became crucial for discovering knowledge nuggets. The need for more storage and processing power led to service-oriented architecture and software/infrastructure-as-a-service analytics business models, democratizing analytics for smaller companies.?

In the 2010s, a paradigm shift in data usage emerged due to widespread Internet use, introducing new data sources like RFID tags, digital energy meters, and social media. Social networking data, characterized by richness and challenges, led to the term Big Data, emphasizing the difficulties posed to computational systems. Hardware and software advancements, such as massively parallel processing and Hadoop with MapReduce, addressed Big Data challenges.?

The future of analytics-related terms remains uncertain, with the time between paradigm shifts in information systems continually shrinking. Despite analytics not being a new concept, its recent popularity explosion, fueled by Big Data and intuitive software tools, has made data-driven insights more accessible. Leveraging analytics presents significant opportunities for businesses to enhance decision-making, increase revenue, reduce costs, improve products, enhance customer experience, combat fraud, and engage customers through targeted customization.?

With global competition intensifying, companies are increasingly equipping their employees with business analytics knowledge to enhance effectiveness and efficiency in decision-making processes. The subsequent section introduces a framework for Business Intelligence (BI), acknowledging its historical significance, while recognizing the evolving landscape toward analytics and data science.?

The term "analytics" has largely replaced individual components of computerized decision support technologies, including Business Intelligence (BI). Analytics, defined as the process of developing actionable decisions based on historical data insights, combines information, computer technology, management science techniques, and statistics to solve real problems. The Institute for Operations Research and Management Science (INFORMS) identifies three levels of analytics: descriptive, predictive, and prescriptive.?

Justifying and prioritizing numerous BI applications poses challenges due to intangible benefits, necessitating knowledge from similar applications and case studies. Security and privacy protection, especially for strategic data, are critical in BI development. Integration of BI applications with various systems, databases, legacy systems, enterprise systems, and e-commerce, along with connectivity to the Internet and business partners' information systems, requires careful consideration.?

SAS Institute Inc. proposed eight levels of analytics, ranging from standardized reports to statistical analysis, forecasting, and decision-making under certain circumstances. INFORMS' three levels of analytics—descriptive, predictive, and prescriptive—depict an interconnected progression, suggesting that one type of analytics application leads to another. The graphical representation in Figure 1.11 illustrates these three levels, emphasizing their somewhat independent yet overlapping nature.?

Big Data analytics is a crucial aspect of analytics and data science, explored in Chapter 7. The challenge lies in the pace of data generation, surpassing computer capabilities for efficient analysis. Big Data refers to data of diverse forms—structured, unstructured, in streams—that cannot be stored in a single unit. Sources include clickstreams, social media posts, and sensor data. Google, facing the daunting task of indexing billions of web pages, successfully employed Big Data analytical techniques.?

Managing data at this scale involves two aspects: storage and processing. The Hadoop Distributed File System (HDFS) emerged as a solution, storing data in chunks on different network-connected machines. To analyze vast data, the MapReduce programming paradigm was introduced, pushing computation to the data. Initially developed by Google, it became Hadoop MapReduce. Companies like HortonWorks, Cloudera, and Teradata Aster offer services in this space.?

Big Data evolved as applications emerged, adding velocity and variety to the equation. Examples include algorithmic trading for rapid data processing in microseconds and sentiment analysis using diverse data sources for gauging sentiments. Today, Big Data encompasses large data with volume, velocity, and variety characteristics. Application Case 1.6 illustrates Big Data analytics in the energy industry, explored further in Chapter 7.?

?

The analytics ecosystem comprises eleven sectors, grouped into technology providers (outer petals), analytics accelerators (inner petals), and analytics user organizations (core). This metaphorical "flower" illustrates the interconnected nature of these components. Technology providers generate revenue by offering solutions and training to analytics users. Accelerators collaborate with both technology providers and users, while the core consists of user organizations driving the industry.?

The outer six petals represent technology providers, and the inner petals signify analytics accelerators. The core of the ecosystem consists of analytics user organizations, the driving force behind every industry cluster. The metaphor emphasizes the symbiotic growth and decline of these components, akin to a living organism.?

The sectors include various players with unique offerings, showcasing where analytics talent can be utilized or recruited. It's essential to note that a company's mention in one sector does not limit its activities, as many operate across multiple sectors. Movement within the field is possible both horizontally and vertically.?

C H A P T E R 8?

Future Trends, Privacy and Managerial Considerations in Analytics?

This chapter explores emerging technologies with significant impacts on business intelligence (BI) applications. The evolving terms in the dynamic field of analytics include BI, analytics, data science, machine learning, artificial intelligence (AI), cognitive computing, Big Data, among others. Instead of delving into subtle differences, the focus is on viewing these technologies as a unified constellation. The chapter examines realized and impending trends in analytics, discussing emerging technologies, their current applications, and their organizational, personal, legal, ethical, and societal implications for managers and professionals in analytics.?

Siemens, a global leader in energy-efficient technologies, aims to leverage Big Data analytics and the Internet of Things (IoT) to predict train component faults in advance. With an annual revenue of $93 billion, Siemens seeks to transition from reactive and preventive maintenance to predictive maintenance for its trains. Equipping trains with sensors, Siemens collects and analyzes data in near real-time. The Teradata Unified Data Architecture, incorporating Hadoop, Teradata Aster, and the Teradata Data Warehouse, processes data from tens of thousands of sensors. Machine learning helps identify genuine part failures, while historical and diagnostic data, weather information, and supply chain data enhance accuracy. Siemens' approach has led to improved train reliability, reduced delays, and increased ridership for train operators worldwide, including RENFE in Spain and Bangkok's trains.?

?

The Internet of Things (IoT) involves connecting physical devices to the Internet, allowing them to collect and transmit data. This data is then analyzed using various analytics techniques for remote monitoring and fault prediction. Examples include self-driving cars, fitness trackers, and Smartbin's sensor-equipped trash containers. IoT applications range from automatic trash collection to a refrigerator ordering groceries. Juniper Research predicts over 38 billion connected things by 2020. The growth of IoT is attributed to smaller, more affordable, and powerful hardware, cost reductions in data processing and mobile devices, and the availability of sophisticated BI tools. The industry is estimated to spend between nearly $6 trillion to $11 trillion on IoT solutions by 2020. Two application cases illustrate IoT's diverse uses across different industries and technology platforms.?

?

Cloud computing is an emerging technology trend defined by the NIST as a model providing convenient, on-demand network access to configurable computing resources over the Internet. This encompasses various services, including utility computing, grid computing, and software as a service (SaaS). Examples like web-based email, social networking sites, and Google Docs illustrate cloud applications. Amazon.com's web services showcase a business example, allowing other companies to leverage its technology infrastructure through a pay-as-you-go model. Major players in cloud computing include Salesforce.com, IBM (Bluemix), Microsoft (Azure), and Google. This trend has influenced new offerings in business intelligence (BI), enabling organizations to scale their data warehouses and pay based on usage. The intersection of cloud computing and BI/business analytics is continually evolving, providing scalable solutions for end users. This section explores a key emerging trend in analytics—incorporating location data. The classification of location-based analytic applications in Figure 8.6 distinguishes between those utilizing static geospatial data and those capitalizing on the dynamic location data generated by contemporary devices. The discussion begins with organizational analytics applications, emphasizing improved decision-making in operational management, customer targeting, and promotions. Subsequently, the focus shifts to consumer-centric analytics applications leveraging location data.?

As data science, analytics, cognitive computing, and AI grow in reach and pervasiveness, everyone is affected by these applications. Just because something is doable through technology, does not make it appropriate, legal, or ethical. Data science professionals and managers have to be very aware of these concerns. Several important legal, privacy, and ethical issues are related to analytics. Here we provide only representative examples and sources. Popular media is usually quite keen to report on such breaches of legal and ethical behavior, so this is one section where you may find even more recent examples online. As pointed out in Chapter 1, our goal here is only to give you an exposure to these issues. Your class teams should identify more recent cases and discuss those. Analytic systems play a crucial role in the ongoing information and knowledge revolution, constituting a rapid and pervasive cultural transformation. Unlike previous revolutions, such as the Industrial Revolution, this current upheaval is swift and omnipresent, affecting various aspects of daily life. The impact of analytics intertwines with other computerized systems, making it challenging to isolate. The consequences of analytics reach both micro and macro levels, influencing individuals, job roles, departmental structures, and even entire industries and societies. The accelerating growth in analytics, AI, and cognitive computing is poised to shape the future of organizations, leaving significant impacts on organizational, individual, and societal levels. The role of a data scientist, closely linked with Big Data, has swiftly become one of the most sought-after positions in the job market. Termed "the sexiest job of the 21st century" by Davenport and Patil in a Harvard Business Review article, data scientists are primarily recognized for their coding proficiency, a skill expected to remain crucial. Beyond technical abilities, effective communication and storytelling with data are enduring skills that data scientists need to possess. Their role involves utilizing a blend of business and technical skills to explore Big Data, aiming to enhance existing business analytics practices and facilitate better decision-making for new business opportunities. Unlike traditional business analysts, data scientists are characterized by their curiosity and inclination to explore new possibilities, often drawing connections between seemingly unrelated fields to derive innovative solutions. This distinctive approach sets data scientists apart in their pursuit of uncovering insights and addressing complex problems.?

In conclusion, data scientists play a pivotal role in today's rapidly evolving landscape, particularly in the realm of Big Data. Acknowledged as one of the most in-demand professions, their significance is underscored by their ability to navigate complex data sets through proficient coding skills. Described as the "sexiest job of the 21st century," data scientists are not only technically adept but also excel in the art of effective communication and storytelling with data. Their unique approach involves a deep curiosity and a propensity for exploring novel possibilities, setting them apart from traditional business analysts. As organizations increasingly rely on data-driven insights, the enduring and multifaceted skills of data scientists position them as key drivers of innovation, uncovering valuable patterns and solutions in the vast sea of information.?