Sunday Dec 29, 2024
Thursday, 6 October 2011 00:00 - - {{hitsCtrl.values.hits}}
The ‘Big Data’ phenomenon seems to have appeared from almost nowhere, but in reality, some aspects of this emerging trend are not new. We have known about the exponential growth of data volumes for some time and according to a recent IDC Digital Universe study published in June 2011, the amount of information created and replicated will surpass 1.9 zettabytes (1.8 trillion gigabytes) in 2011, growing by almost nine times in just five years.
Other aspects of ‘Big Data’ are indeed new.
The variety of the data sources is growing at a rapid rate, particularly as we move into the semi-structured and unstructured realm (e.g. social media interactions, rich media files and geospatial information). The other emerging factor that organisations need to contend with is the increased velocity at which data is being generated (e.g. real-time sensor data feeds from smart meters).
“These new aspects of ‘Big Data’ are creating unprecedented levels of complexity for IT executives, particularly as they realise that these massive data sets cannot be processed, managed and analysed using traditional databases and architectures,” says Philip Carter, Associate Vice President at IDC Asia/Pacific. “What is becoming clearer is that the real value from ‘Big Data’ will be derived from the high-end analytics, predominantly using data mining, statistics, optimisation and forecasting type of capabilities to proactively turn this data into intelligence to drive business benefits and better decision making capabilities.”
In line with this trend, as businesses in Asia invest to drive growth in emerging markets, they are harnessing analytics-led solutions to gain better customer insights, and manage risk and financial metrics more effectively while striving for unique market differentiation. In a February 2011 C-suite barometer survey drawing over 1000 responses from CIOs and LoBs across Asia/Pacific, IDC found that business analytics ranked as the top rated technology that would allow organisations to gain significant competitive advantage in the year ahead. In addition, in an IDC June 2011 survey of over 1300 CIOs and IT decision makers across Asia/Pacific (excluding Japan) or APEJ, data management and analytics ranked as the top business priority for organisations in the region.
However, the approach to business analytics in the era of ‘Big Data’ will be significantly different to the traditional approach. “For example, one of the key differences between traditional analytics and what we are dealing with in terms of the ‘Big Data’ era is that we are gathering data that we may or may not need. From an analysis perspective, this means ‘we don’t know what we don’t know’. To run an analysis on ‘Big Data’, the variables and models are likely to be entirely new. Therefore, a different infrastructure strategy and perhaps most importantly, new skill sets, are required,” adds Philip.
To cope with the challenges ‘Big Data’ poses, organisations must begin looking at deploying not only the applications traditionally used for Business Analytics (BA), but also the supporting architecture in order to scale efficiently. IDC recommends looking at cloud bursting, the deployment of analytical appliances, and creating truly scalable enterprise architectures that leverage the attributes of high performance computing.
This approach should also allow for the deployment of new technologies and frameworks such as Hadoop to assist with the analysis of large pools of disparate, unstructured data. This will require new technical skill sets – particularly around emerging technologies like Hadoop, Map Reduce and Key Value Stores – as well as a revised approach to the role of the business analyst. The next generation business analyst will be more akin to “data scientists”.
These individuals will have strong statistical skills and will be able to extract information from large datasets and present value to non-analytical experts. They will also have the unique skill of understanding the new algorithms and analytical models that will have the most significant business impact in the short term.