Identify A Big Data Problem
How big is Big Data Project?
Many companies have a data sets of hundreds, thousands and even lakhs of GB.That’s a lot of data, but rarely meets the litmus test for true big data problem. Big data Projects have large, diverse, distributed and mixed data sets characteristics:
- Do you have Perabyte of data?(1048576 GB’s). If “no”, you don’t have a big data problem.
- Do you really need responses in real time? If “no”, you don’t have a big data problem.
- Does it take long time to calculate results? (more than 8 hours or days) If “no”, you don’t have a big data problem.
Are the answers provided by big data really a worth investment of hundreds and thousands of dollars into the business. If “YES”, your business right for utilizing big data.
Characteristics of Organizations that could use BIG DATA
- The organizations producing data are big. Think of Government Agencies and Fortune 1000.
- Producing data is an integral part of their daily activities.
- They May have a massive number of web users that produce information.
- They produce a large amount of information on Biomedical, Scientific and Engineering Research.
- They use equipment that create a new stream of data continuously.
- They need solutions in real time to solve critical tasks that can be solved using big data technologies (also consider checking out this perfect parcel of information for data science degree).
- Their core business is to produce data and sell the data sets around the data they created.
- They have sponsors supporting and investing the big data initiative overtime.
How do a Big Data Problems Look like?
Much of the Big Data buzz is centered on mining social data like Facebook info, Twitter feeds etc. It becomes a market researcher’s gold-mine after some really smart people come up with the right models to make sense of all non categorical information. So now, you can see how a certain topic or brand is tending. Its good. But what will you do with the answers provided by Big Data?
Lets say you know your Market Demographic cold. What would you offer a particular customer as they browse your online storefront. That would translate to dollars. So, how would you know all the information about customers?Its log file events. This is where you track each step of user’s experience on a website for example.
How does the customer log file information help you? Having lots of log files on their own doesn't help you to make decisions. Smart people should take an action on the data and really do something with it. The immediate next step is to create algorithms that allow to predict what customer want to buy on the current visit on the website.
Why is there a so much hype on big data?
- Because you are solving a problem of Big Data that consumers think its valuable. Then you will have a fantastic Competitive advantage
- Because a lots of data is being produced by machines, individuals, Websites, and other sources. We need a tool to make use of that right?
- Because Engineers, Scientists and professors can use big data to solve complex problems and ever evolving theories (also consider checking out this career guide for data science jobs).
Big data Problems are Fantastic, Challenging and always Changing.
Author : Lisa Findlay