Introduction to Text Mining Part 1

Qualitative data can have a lot of insight into the mind of the customer, but it isn't the most useful tool for analysis. While there are some ways to code qualitative data, the reality is that some people don't provide complete thoughts at all, others are unable to put their thoughts into words, and still others find that their opinions are misunderstood. But from a customer satisfaction perspective, qualitative data can be surprisingly beneficial. There are many anecdotal examples of businesses that learn a great deal from a single qualitative answer, and come up with great ideas straight from those that were willing to provide them.

Still, unless you come up with an easy way to analyze your qualitative data, or you have many researchers on staff that all know exactly how to code it, it's difficult to get much meaning from your responses. That's why some companies turn to text mining. Text mining involves a complete program – one that can be built for you or purchased from various sources on the internet.

The program goes through all of the data and mines it for recurring words, phrases, or themes. It can't tell you what the data means (it is, after all, a computer program), but it can present to you a chart or graph that shows you what information has recurred at how many intervals in the survey. In a way, text mining is a form of quick analysis.

By running qualitative answers through the program, you are instantly able to receive a graph that gives you something to focus on, and an idea of something that may be of interest to you and the customer.

Text mining has several possible benefits to your company: Missed Information Because text mining looks for words rather than meaning, you may be able to locate information you hadn't noticed or considered prior to the mining taking place. By taking the human element out of analysis, you find an objective way to look at responses and see if there is anything you need to explore further. Initially Quick There is value to going over each qualitative answer individually, but that also takes a lot of time and effort.

If you want to analyze results right away or get a glimpse of what you may need to look into more eventually, text mining is fast enough to provide you with near instant results. We'll continue in the next article.