Подготовка к путешествию в мир Больших Данных

Today, leading organizations run petabyte-scale data warehouses, but traditional analytic platforms are not suited for the scale and complexity of such big interaction data as social media content, sensor and machine information, call detail records (CDRs), and Web logs. Organizations need a new approach to harness unstructured and multistructured information and to combine it with ever-growing volumes of big transaction data (from data warehouses, ERP applications, and OLTP systems).

With the right approach, big data promises to help organizations improve operational efficiency, combat fraud and customer churn, enhance products and services, and increase insights and profitability. Capitalizing on the business potential of big data is of keen interest in most industries:

• Financial services: Detect fraud; enable more immediate and precise marketing
• Telecommunications: Leverage CDRs to optimize networks; reduce customer churn
• Manufacturing: Streamline production; strengthen quality control
• Healthcare: Generate insights from clinical data; improve patient outcomes
• Public sector: Enhance public safety, disaster preparedness, and research
• Consumer industries: Deepen customer engagement with social media analytics

Size Up Your Analytics Maturity

The first step in a big data journey is to understand the maturity of your organization from an analytics perspective. The business intelligence (BI) infrastructure and processes you have in place will influence how rapidly your business and IT teams can marshal the resources needed to capitalize on the big data opportunity. The research and advisory firm Bersin & Associates offers a good analytics maturity model to benchmark big data readiness (see Figure 1).

• Level 1: Reactive operational reporting with classic dashboards is geared to measure how an organization has performed; these lowest-level analytic processes can consume weeks and lack complete, detailed, timely data.
• Level 2: Proactive advanced reporting is a step up, supporting improved decision making with drill-down and segmentation of more statistically robust information, yet is still a rear-view-mirror look at performance.
• Level 3: Strategic analytics give organizations a forward-looking view based on advanced modeling and deeper segmentation to produce actionable insights and identify opportunities and causes of issues.
• Level 4: The most advanced level of predictive analytics uses predictive modeling and pattern discovery to forecast future conditions and mitigate risk, based on complete structured and unstructured data.

In consultation with many hundreds of enterprises across industries, Cognizant believes that most organizations are at Level 2 of analytics maturity. At the same time, the majority considers advancing its BI practice and data integration capabilities a high priority as big data continues to emerge.

Many enterprises are in early phases of testing and production with such big data technologies as the opensource Hadoop/MapReduce data processing framework. An Informatica® survey of 589 IT and business professionals in North America, Europe, and Asia Pacific in spring 2012 found that 70 percent of respondents had big data projects in planning or production.

By sizing up your big data readiness, you position yourself to start building a foundation for leveraging new and larger data sets. It can be risky, however, to assess maturity or build a foundation in a vacuum. With its worldwide EIM practice, Cognizant has extensive experience in helping organizations benchmark their readiness and navigate what can be a confusing maze of technologies and vendor claims about big data objectives.


Source:  informatica.com