Big Data holds great promise for enterprises of all sizes. It can bring insights that help the business drive revenue and also understand gaps in service and products.
Here are some myths about data:
1. Big data is new
Huge cross references of every single word used in the Bible,called “Concordances”,were in use by scholar monks for centuries well before the first databases.
2. Big data is made for Big business
Enterprises of all sizes are able to now leverage big data analytics thanks to recent improvement in cloud and data management technology.
3. Bigger Data is Better
Quality of data wins over quantity of data.What to use is often more relevant than how much to use.
4. Our data is so messed up we can’t possibly master big data
Advanced data quality,master data management,and data governance tools have made it easier to clean up the enterprise data mess.
5. Every Problem is a big data Problem
If you are matching a couple fields against a couple of conditions across a couple of gigabytes,it isn’t really a big data problem. Don’t treat every analytics need as a big data effort.
6. Big Data Application require little or no performance tuning
Big Data application require regular tuning of the analytical and statistical models are more and more data and variables are added.
7. Big Data is a Magic 8 Ball
Big Data may not tell you everything. A lot depends on the right questions and the right data for it to work.
8. Big Data is only unstructured data
Big Data does not have to be unstructured. Even voluminous structure data may classified as Big Data because of its sheer volume.
9. Machine Learning is a concept related to Big Data
The idea underlying machine learning is “ using data to model an underlying process”. Machine learning algorithms can,however,provide valuable insights when used in conjunction with Big Data.
10. Big Data analytics will not require supervision by humans
The unsupervised does not mean that these algorithms run by themselves without human supervision. An analyst who is training an unsupervised learning model has to exercise a similar kind of modelling discipline as the one who is training a supervised model.