The Human Side of Big Data and High-Performance Analytics - PDF

Please download to get full document.

View again

of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report

Taxes & Accounting


Views: 8 | Pages: 13

Extension: PDF | Download: 0

Related documents
Research Report The Human Side of Big Data and High-Performance Analytics August 2012 Authored by: Thomas H. Davenport Executive Summary In order to succeed with big data and high-performance analytics,
Research Report The Human Side of Big Data and High-Performance Analytics August 2012 Authored by: Thomas H. Davenport Executive Summary In order to succeed with big data and high-performance analytics, organizations require not only new technologies, but also a new set of human capabilities. A key component of these capabilities are data scientists hybrids of analytical and computational skills, typically with science backgrounds. Data scientists are motivated not just to support internal decisions with analytics, but to create new products and processes for customers. In addition, big data and high-performance analytics require new approaches to deciding and acting on the part of executives and decision-makers. The concept of big data burst onto the technology and business scene in 2010, and since then has excited many executives with its potential to transform businesses and organizations. The concept refers to data that is either too voluminous, too unstructured, or from too many diverse sources to be managed and analyzed through traditional means. The definition is clearly a relative one that will change over time. At the moment, however, Too voluminous typically means databases or data flows in petabytes (1000 terabytes); Google, for example, processes about 24 petabytes per day of data. Too unstructured generally means that the data doesn t initially come in numeric form, and isn t easily put into the traditional rows and columns of conventional databases. Examples of big data include a massive amount of online information, including clickstream data from the web, and social media content (tweets, blogs, wall postings, etc.). Big data also incorporates video data from retail and crime/intelligence environments, location data from mobile devices, and data from the rendering of video footage. It includes voice data from call centers or intelligence interventions. In the life sciences, it includes genomic and proteomic data from biological research and medicine. Many IT vendors and solutions providers, and some of their customers, are treating the term as just another buzzword for managing and analyzing data to better understand the business. But there is more than vendor hype; there are considerable business benefits from being able to analyze big data on a consistent basis. And given the pervasiveness of devices that generate data from smartphones to RFID chips to sensors in manufacturing technologies and manufactured products we can expect big data in almost every industry. Companies that excel at big data will be able to employ these technologies with business benefit, and to harness the Internet of things. They will produce new products and services based on data. They will ultimately be able to understand their business environment at the The Human Side of Big Data and High-Performance Analytics August 2012 p. 2 most granular level, and adapt to it rapidly. They ll be able to differentiate commodity products and services by incorporating monitoring and analysis of usage patterns. And in the life sciences, of course, effective use of big data can yield cures to the most threatening of diseases. Companies with more traditional types of structured data can also benefit from recent technology developments. Advances in parallel computing, in-database analytical processing, and in-memory software technologies now make it possible to dramatically accelerate the pace with which analytical insights can be delivered. Often called high-performance analytics (HPA), these tools can reduce the cycle time for complex and large-scale analytical calculations from hours or even days to minutes or seconds. The department store chain Macy s, for example, has been able to reduce the time to optimize pricing of its 73 million items for sale from over 27 hours to just over 1 hour. HPA obviously makes it possible to reprice items much more frequently to adapt to changing conditions in the retail marketplace. But prospering from big data is not simply a matter of employing new technologies. To achieve the benefits from big data and high-performance analytics, firms will need to make some adjustments to their capabilities, even if they are already experienced users of analytics. Big data and high-performance analytics environments are clearly different from traditional data analysis environments in many ways. In this white paper, however, the focus is on the humanrelated differences. A key focus is on the data scientists who do this sort of work, and the implications for executives and managers will also be considered. See the sidebar About the Research for more details on the research approach employed. Sidebar: About the Research In order to learn more about data scientists and their activities, I interviewed (primarily by telephone) 30 people in the first half of 2012 who described themselves as data scientists, and several more managers to whom data scientists reported. In all but a couple of cases, it was clear that the individuals interviewed had data scientist-oriented backgrounds and were actually performing data science activities. The data scientists came from a wide variety of industries, including online data and services (8 of 30, or 27%), IT vendors (20%), professional or IT services (13%), health care (13%), manufacturing (7%), media (7%), and several industries with only one respondent. I asked about a variety of topics, including their backgrounds, motivations for entering work in big data, current activities, reporting and working relationships, and views on what makes for effective data science teams. The Human Side of Big Data and High-Performance Analytics August 2012 p. 3 The Rise of Data Scientists While there has always been a need for analytical professionals to support an organization s analytical capabilities, with big data the requirements for support personnel are a bit different. The fact that dealing with the data itself obtaining, extracting, manipulating, and structuring it is an important prerequisite to doing any analysis means that the people who work with big data need substantial IT skills. The role of analytical professional in big data environments has been christened as the data scientist. Data scientists understand analytics, but they also are quite skilled at exploring and exploiting information technology. Data scientists work in a variety of organizations, from big data startups (37% of sample interviewed) to large, established companies like GE, Intuit, and EMC (63%). GE, for example, expects to hire over 400 computer and data scientists at its new Global Software and Analytics Center in the San Francisco Bay Area. While some of the GE data scientists will work on traditionally data-intensive problems in financial services, logistics, and health care, a significant portion of their focus will be on big data for industrial products, such as locomotives, turbines, jet engines, and large energy generation facilities. Data scientists have somewhat different roles from traditional quantitative analysts. Whereas traditional analysts typically use analytics on internally generated data to support internal decisions, the focus of many data scientists is on customer-facing products and processes, where they help to generate products, features, and value-adding services. At the business social network LinkedIn, for example, data scientists developed the People You May Know and Jobs You May Be Interested In features of the site, among others. GE is already using data science to optimize the service contracts and maintenance intervals for industrial products. Google, of course, uses its roughly 600 data scientists to refine its core search and ad-serving algorithms. Zynga uses data scientists to target games and game-related products to customers. The testing firm Kaplan uses its data scientists to begin advising customers on effective learning and test-preparation strategies. In health care big data firms, data scientists try to discover the most effective treatments for different diseases. Given this focus, data scientists are most likely to be in product development or marketing organizations (just over half of my sample) Some also work for Chief Technology Officer (CTO) organizations. Some firms have established new positions to which data scientists report, including a Senior Vice President of Big Data, Social Design, and Marketing (at Intuit), and a Chief Digital Officer (at Kaplan). Data scientists are not likely to work closely with corporate IT organizations or Chief Information Officers, though stronger relationships with those functions The Human Side of Big Data and High-Performance Analytics August 2012 p. 4 are sometimes found in health care big data firms. Several did mention, however, that they relied on traditional business intelligence functions for help with reporting of the results from big data analyses. Those who report to CTOs are likely to work on tools that make data science easier and more productive. Because the tools for managing large volumes of unstructured data are still evolving rapidly (the widely used Hadoop framework for distributed file system processing only went into production at Yahoo in 2008, for example), some data scientists are focused on creating and improving tools. Yahoo thus far hasn t used data as a product, but rather built an infrastructure for the rest of the world to use. This has also been a primary activity, for example, of Facebook s data team, which has created the language Hive for programming Hadoop projects. Some firms, such as Yahoo, Facebook, and Microsoft, also support academic-style research and publication by their data scientists. This may be useful for retention of the data scientists, although the business value of the activity is less clear. Data scientists who focus on HPA applications don t necessarily need to understand how to process unstructured data, but they do need to understand how analytical work can be divided across multiple parallel servers. They may also serve as process consultants, helping their organizations to speed up business processes based on the dramatic reductions in time required to produce extensive analytical results. They should be able to explore a variety of ways to use the extra time from HPA to refine their models. In addition, as I will discuss in greater detail below, they need to try to accelerate decision speeds to match the much faster cycle times of data analysis. Data Scientist Skills Data scientists require technical, business, analytical, and relationship skills. From a technical standpoint, many have advanced computer science degrees, or advanced degrees (often Ph.D.s 57% of my sample) in fields such as physics, biology, or social sciences that require a lot of computer work. 90% of the data scientists I interviewed had at least one degree in a scientific or technical field. Almost all had strong computational skills. They re not just programmers; many refer to data scientists computational skills as hacking technology bending it to do their bidding in unusual ways. The Human Side of Big Data and High-Performance Analytics August 2012 p. 5 LinkedIn, the social networking site targeted at business professionals, has a substantial data science team, and has had considerable success developing data-driven products and features. A job description for the data scientist role at LinkedIn describes the desired technology traits: Are you someone who solves hard problems by creatively obtaining, scrubbing, exploring, modeling and interpreting big data? Do you know enough about information retrieval, machine learning, and statistics to be dangerous? Are you a hands-on implementer, ready to learn new languages and technologies to turn your ideas into solutions used by tens of millions of people around the world? As the job description suggests, data scientists tend to address a few key technologies. They include: Hadoop, MapReduce, and the related ecosystem of distributed file system tools; Programming languages such as Python, Java, Pig and Hive; Machine learning; Nontraditional database tools such as Vertica and MongoDB; Natural language processing; Statistical tools. Overall, data scientists particularly those in startup firms seem to have a strong preference for open-source tools and technologies. In addition to these technical skills, data scientists also need the attributes previously necessary for analytical professionals. These include mathematical and statistical skills, business acumen, and the ability to communicate effectively with customers, product managers, and decisionmakers. Several data scientists commented that relationships are important to success in the field. This was also true with traditional quantitative analysts, but the relationships there were with internal business decision-makers, rather than product and process managers. Of course, the combination of these skills is difficult to find in one person. Some companies, reflecting this problem, have created data science teams that together embody this collection of skills. Each individual member of the team may have only some of the necessary skills, but assuming they work closely together, they can do all the necessary activities. The Human Side of Big Data and High-Performance Analytics August 2012 p. 6 Recruiting or Creating Data Scientists Adding the hard-core technical requirement to traditional analytical skills makes it even more difficult to find such individuals. At a recent gathering in Silicon Valley on big data, the consensus of the attending experts was that finding qualified data scientists in sufficient numbers will be the greatest constraint on the field. And in a recent Economist Intelligence Unit survey of 600 global executives, 54% of North American respondents said finding the right people with the right skills is the No. 1 obstacle to launching a successful big data project. Where can an organization find data scientists? There are few if any academic programs in the area, although several are being designed now. Some existing master s degree programs in analytics, such as that at North Carolina State, are including some big data training in their curricula (such as Hadoop programming and dealing with unstructured data). Most organizations, however, must recruit and hire individuals from other backgrounds who have skills related to data science. For example, George Roumeliotis, the head of a data science team at Intuit in Silicon Valley (and himself a Ph.D. in astrophysics), doesn t hire purely on the basis of statistical or analytical capabilities; instead, he needs people who can develop prototypes in a mainstream programming language such as Java. He says he has given up trying to recruit anyone with industry experience they just don t exist so he primarily recruits directly out of schools (Ph.D. programs). Roumeliotis seeks both a skillset a solid foundation in math, statistics, probability, and computer science and a mindset a feel for business issues and an empathy for customers. Then, he says, he builds on that with a mixture of on-the-job training and an occasional course in a particular technology. Given the difficulty of finding and keeping data scientists, one would think that a good strategy would involve hiring them as consultants. Yet most firms that are aggressively pursuing big data projects seem to want to employ their own data scientists (perhaps because they are worried about turning over their important data assets to outside firms), and most consulting firms have yet to assemble them in large numbers. Firms such as Accenture, Deloitte, and IBM Global Services do have data scientists on staff (some call them management scientists or decision scientists ), but they are only in the early stages of leading big data projects for large-firm clients. Thus far they are primarily applying their skills to more conventional quantitative analysis problems. As client demand for big data work builds, they will no doubt offer more data scientists-for-hire. The Human Side of Big Data and High-Performance Analytics August 2012 p. 7 There are a variety of other approaches in use to develop and hire data scientists. EMC, for example, has determined that the availability of data scientists will be an important gating factor in its own big data efforts and those of its customers. So it has created a Data Science and Data Analytics training program for its employees and customers. EMC has already begun to put graduates of the program to work on internal big data projects, and has also made the course materials available to universities. One data scientist has come up with a creative approach to training new data scientists. The Insight Data Science Fellows Program, started by Jake Klamka (a high energy physicist by background) takes scientists for 6 weeks and teaches them the skills to be a data scientist. The program includes mentoring by local companies with big data challenges (e.g., Facebook, Twitter, Google, LinkedIn, etc). I originally was aiming for 10 Fellows. I had over 200 applicants and accepted 30 of them, says Klamka. He goes on, The demand from companies has been phenomenal; they just can't get this kind of high quality talent. Venture capital firms are also entering the data science game. In order to help the demand by companies in their portfolio, Greylock Partners, an early stage venture firm that has backed companies like Facebook, LinkedIn, Palo Alto Networks, and Workday, has a built a recruiting team that focuses in part on data scientists. Dan Portillo, who leads the team, says, The demand of data scientists is at an all-time high from our later stage companies. Once they have data, they really need people who can manage it and find insights in it. The traditional backgrounds of people you saw years ago, just don't cut it these days. Once they are hired or created, companies may also face issues in retaining data scientists. Several of those I interviewed had changed jobs several times in the past year. One commented, After about a year it often becomes clear that there is nothing left for me to do. Another noted, Data scientists receive lots of job offers sometimes I get two or three calls a week from headhunters. It s not surprising that with so much opportunity there is a lot of movement. In order to hold on to the data scientists in your organization, you need to offer them a combination of autonomy (to be able to make an impact with their work), interesting and useful data, and a state-of-the-art technical environment all in addition to a lucrative compensation package. What Motivates Data Scientists? If you re interested in recruiting and retaining data scientists for big data work, it s important to know what motivates them. In my interviews with data scientists, the same motivational issues The Human Side of Big Data and High-Performance Analytics August 2012 p. 8 appeared frequently. Data scientists want to use data to have a substantial impact on the world. They view this as a unique period in history in which there are huge datasets and very powerful tools. As Amy Heinike, a prominent data scientist at the startup Quid in San Francisco put it: If you have access to the data and the tools, you can already find out some really cool stuff, but we are just scratching the surface. What inspires me is the opportunity to create something really interesting. Could something be important, have impact, or reach a lot of people? I am also interested in how to include data scientists within a diverse engineering team and company, and in combining the diverse skills that make up an effective data science team. So when I evaluate an opportunity, I look for a rich dataset that the company has to work with, or an important question for which we might be able to find data. And I want to make sure that there are the resources and senior management support available to support the d
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!