With answers to this (short) list of initial questions, we will be ready to start the planning process that will lead the business to exploit Big Data. Meantime, we need to keep the lights on with what we have today.
Big Data and analytics have often been mentioned in the same breath, particularly with reference to predictive analytics which seems to be the "holy grail" of data analysis. Are we there yet?
Predictive analytics have been with us for some time. The new value that Big Data brings is both accuracy and time. As an example, it has long been possible to identify which audience has been predicted for a particular TV show or print magazine. If the advertising is as interesting to the audience as the media content then the predictive analysis was good. TV audiences and print advertising has a very slow decision cycle. TV content is created sometime years ahead, and print magazine content is prepared some months in advance.
In the case of TV advertising, road traffic patterns, or impending component failure in turbine engines, we have used statistical probabilities to predict outcomes, and in all cases we have accepted that there is an acceptable median with some outliers.
Big Data and fast data together are beginning to give us totally new options. In TV advertising, where we can "know" the audience based on known subscribers to IPTV services or specific individuals and locations based on cellular viewing patterns. In road traffic we can sense the actual movement and automate control and in turbine engines we can sense the performance of every component and observe insipient failure. This all gives us better predictive and reduce waste.
Pivotal provide both Big Data "deep" analytics with our high speed Pivotal Hawq solution for rapid analysis of very large data sets, using Pivotal Data fabric platforms. We also offer our Pivotal Gemfire solution for analysis of very large "high velocity" data streams. By putting both Pivotal Hawq and Pivotal Gemfire together it is possible to establish fast and accurate predictive analytics across a wide range of business use cases from real-time risk management to active device failure in mechanical components.
From the telco's perspective, what challenges are they struggling with in formulating their Big Data strategies? Or have they taken giant strides in that direction to really benefit from big data and analytics?
The transition of the global voice communications market from fixed to mobile is, in many markets, nearing completion. The same transition is well underway for data communications. One of the consequences of this transition is the transition from shared equipment to personal connected devices. Cellular Network Service Providers have remained slow to exploit the data available to them. Complexity of national privacy law, coupled with stringent government control over telecommunications company licensing have created a tendency to play safe.
Sign up for CIO Asia eNewsletters.