What really is big data? Big data encompasses extremely large datasets that can be analyzed to reveal more in-depth insights, patterns, trends, and even help predict future outcomes. But what actually makes up these “extremely large datasets” can be much more exhaustive, and understanding them can significantly improve our overall knowledge of big data and how to use it.
As we’ve started to mention, big data is just data. And the following types of big data can be used to define any data in today’s world. But the goal of understanding the different types of data is to help determine how they might be used together to provide the answers to the questions researchers and market analysts may be asking about data.
Types of Data
First and foremost, big data can be defined based on its structure. The structure of data depends on how organizable it is. In other words, whether it can be formatted into tables of rows and columns. There are three types of big data when defining it by the structure:
- Structured: Data that is structured is often already stored in a database or other data management platform, and it can be easily accessed and processed to provide an ordered output.
- Unstructured: Usually larger datasets—the majority of big data is unstructured, meaning it can’t easily be organized or classified.
- Semi-Structured: As the name implies, semi-structured data isn’t inherently organized at the start, but as it is analyzed or digested it can begin to take on a more structured form.
Both structured and unstructured data can be either human-generated or machine-generated. Human-generated, structured data can be contact information or website form details directly collected from an individual. Human-generated unstructured data can be any form of website activity and social data such as video, audio, or social posts shared by a person.
On the other hand, examples of machine-generated, structured data include GPS tracking, inventory tracking, or transaction data. Unstructured forms of machine-generated data include information gathered through satellite such as images or weather sensory information.
Each of these types of data can be analyzed in many different ways. However, there are certain types of analysis that will serve their own purpose depending on the objectives at hand.
Types of Analysis
There are many reasons to look to big data for insights. Whether it’s combining big data and survey data for detailed audience intelligence or combing through it to predict purchase data, they all fit into four types of analysis:
- Prescriptive Analysis: Data analysis that provides answers to what actions should be taken.
- Predictive Analysis: An analysis of data that can be used to predict what situation or number of situations may results.
- Diagnostic Analysis: Data analysis that provides insight into what happened in the past and why.
- Descriptive Analysis: Data analysis that can be real-time or leveraged to see what is currently happening.
Mapping a research approach to the type of big data needed and the type of analysis can help understand what tools and solutions may be best to bring it all together. Specifically, the type of data and analysis will lead you to the type of big data analytics required. To learn more about how we combine structured big data with survey data to answer audience questions, watch the webinar below.