Hadoop is a data system so big it is like a virtual jumbo where your PC is a flea. One of the developers named it after his kid’s toy elephant so there is no complicated acronym to stumble over. The system is actually conceptually simple. It has loads of storage capacity and an unusual way of processing data. It does not wait for big files to report in to its software. Instead, it takes the processing system to the data.
The next question is what to do with Hadoop. Perhaps the question would be better expressed as, what can we do with a wonderful opportunity that we could not do before. Certainly, Hadoop is not for storing videos when your laptop starts complaining. The interfaces are clumsy and Hadoop belongs in the realm of large organisations that have the money. Here are two examples to illustrate the point.
Hadoop in Healthcare
In the U.S., healthcare generates more than 150 gigabytes of data annually. Within this data there are important clues that online training provider DeZyre believes could lead to these solutions:
- Personalised cancer treatments that relate to how individual genomes cause the disease to mutate uniquely
- Intelligent online analysis of life signs (blood pressure, heart beat, breathing) in remote children’s hospitals treating multiple victims of catastrophes
- Mining of patient information from health records, financial status and payroll data to understand how these variables impact on patient health
- Understanding trends in healthcare claims to empower hospitals and health insurers to increase their competitive advantages.
- New ways to prevent health insurance fraud by correlating it with claims histories, attorney costs and call centre notes.
Hadoop in Retail
The retail industry also generates a vast amount of data, due to consumer volumes and multiple touch points in the delivery funnel. Skillspeed business trainers report the following emerging trends:
- Tracing individual consumers along the marketing trail to determine individual patterns for different demographics and understand consumers better.
- Obtaining access to aggregated consumer feedback regarding advertising campaigns, product launches, competitor tactics and so on.
- Staying with individual consumers as they move through retail outlets and personalising their experience by delivering contextual messages.
- Understanding the routes that virtual shoppers follow, and adding handy popups with useful hints and tips to encourage them on.
- Detecting trends in consumer preferences in order to forecast next season sales and stock up or down accordingly.
Where to From Here?
Big data mining is akin to deep space research in that we are exploring fresh frontiers and discovering new worlds of information. The future is as broad as our imagination.