What is Unstructured Data and Why use it?
Briefly, Unstructured Data is a term that “…refers to information that either does not have a pre-defined data model or is not organized in a pre-defined manner”. But, what does that mean? Picture an excel sheet with labelled columns. This is considered structured data because a computer can understand each cell on an informational level and it can be manipulated in order to create insights. Data sets that aren’t considered structured are free-form text (i.e Facebook comments, user feedback, tweets, etc..), images, and audio to name a few (Find a whole list here).
So why is this relevant? It is extremely relevant because this is the age of data and ~90% of it is unstructured. We are also at a time in technology where processing unstructured data is becoming much more feasible. In fact, that is what we do over at Semeon Analytics: we process unstructured text data such as social media comments and resumes. We make sense of information that only used to be done by humans.
What’s preventing you from analyzing your data properly?
They say that in the knowledge economy, information is our most valuable commodity. But, a study in 2013 revealed that despite the huge volumes of data now globally produced only 0.5% of the data is actually analyzed. So why aren’t we making sense of the data if it’s our most valuable commodity? In the world of big data, there are 4V’s: variety, volume, velocity, and value. Each variable contributes to the usefulness of a companies data. They are also the reason so much data does not get measured.
- It is too varied or not varied enough
- There is too much of it or too little of it
- It is coming too fast or too slow
- Insights would bring little value
These variables are what make data so hard to measure and create actionable insights from large sets of data. Now back to unstructured and structured data. As mentioned above, 90% of data is unstructured. So, not only do companies need to get all these variables correct, they also have to deal with data that can’t be processed by a computer. Fortunately though, with advancements in artificial intelligence, there are a few companies able to process unstructured data on a contextual level. At Semeon, we focus on creating insights from unstructured text data that bring value to a company. Specifically, by using the following techniques:
- Sentiment Analysis
- Concept Clouds
- Timeline Tracking
- Content Classification
- Influencer identification
- Data Visualization
- Intent Analysis