Big Data Basics
What is Big Data Analytics?
Big data analytics helps businesses to get insights from today’s huge data resources. People, organizations, and machines now produce massive amounts of data. Social media, cloud applications, and machine sensor data are just some examples. Big data can be examined to see big data trends, opportunities, and risks, using big data analytics tools.
Big Data Basics
Until recently, data was mostly produced by people working in organizations. The data usually had a specific structure. It was the basis of records for money paid, deliveries made, employees hired, and so on. This data is still vital to businesses. Now, big data concepts mean that data processing must manage:
- High volume (lots of data)
- High velocity (data arriving at high speed)
- High variety (many different data sources and formats)
Big data can be structured, but with high volume, like historical payment transaction data. It can be semi-structured as in XML and other user-defined content. It can also be totally unstructured. Free form text used in social networks is an example.
Using Big Data Analytics
The more data you have, the more chance you have of getting useful insights from it. However, the size of big data usually makes it impossible to use manual or even conventional computing methods (learn more here: big data and Hadoop). Instead, big data analytics is based on:
- Data mining to sift through data to find patterns and relationships
- Statistical algorithms to build models and predict outcomes
- Machine learning to handle changing and new data, to adapt and enrich models
- Text analytics and natural language processing to analyze free form text and speech
Big data analytics tools can also be grouped like this:
- Descriptive analytics to show what happened
- Diagnostic analytics to explain why something happened
- Predictive analytics to suggest what will likely happen next
- Prescriptive analytics to tell users what to do, to obtain a given result
An Example of Big Data Basics and Analytics in Action
Suppose a company runs big data analytics on its past sales data. It sees that demand has been rising in certain regions for one of its product lines (descriptive). From additional social media and CRM data, it also finds that customers are buying products from this product line to replace a competitor’s product (diagnostic). Adding in marketing data and analytics, the sales potential is forecast for each region where customers might replace their current product (predictive). Recommendations are then made for optimizing advertising and pricing to maximize sales profits (prescriptive).
See it in action: