Text (aka unstructured) data is real, it’s huge, and it’s rapidly building up (both internally – at your company, and externally – on the web). You’ll only see more in the days to come. And guess what? If you are practicing traditional analytics at your company, foraying into unstructured data analysis will probably be one of the most intriguing challenges you’ll ever face. One simple reason: You may not know what to do with it.

Even if you know what to do, the bigger question to ask yourself is, “How do I do it?” We answer this question in our new book, Text Mining and Analysis: Practical Methods, Examples, and Case Studies Using SAS. We provide you with the information you need to get started exploring the realm of unstructured data, so you can learn the fascinating methods you can follow to tweak your text data and churn useful information.

If you are an analyst looking forward to analyzing unstructured data using SAS, you should keep these important things in mind:

1. Your first step should be to establish your primary goal(s) – What do you want to achieve from your unstructured data analysis?

It is absolutely essential for you to think about this when it comes to unstructured data. For data analysis involving structured data, you usually rely on the data stored in your database backyard, whether it is generated from a point-of-sale or another transaction record generating system.

For example, you may wish to explore your call center notes for emerging issues, topics or trends, and group documents based on similarity of the content (discussed in chapters 4 through 6); or you might want to tag documents in your repository with metadata for improving information search and retrieval process (chapter 7). In those cases, you know what the text data is and where you can find it.

In the case of unstructured data, depending on your primary objective, you may not always have data readily available. For example, if you wish to understand public sentiment about your company’s products, services, or even brand image perception (chapter 8), then you may have to start looking for other data sources which can provide you with the necessary data, both internally and externally. In yet some other cases, you may already have a good model (such as churn) that uses only numeric data and you want to improve this model by bringing in text data. No worries – we show you how to do this in chapter 5 and through a case study.

2. You don’t need to know SAS programming to analyze text data using SAS

You heard it right! Most of the SAS Text Analytics tools provide you with an intuitive, user-friendly point-and-click interface to perform analysis and build projects. Throughout the book, we have provided examples with screenshots to show you how to navigate, use, and interpret results in a non-programming paradise mode (our new patented terminology for those of you who don’t like to code).

Note: The only exception is when it comes to extracting complex concepts and context based facts from the text. In such cases, you may be required to write some concepts rules with minimal effort.

3. The proof of the pudding is in the eating.

We discuss the theoretical background of the SAS text mining tools and their functionalities in this book because we wanted to make sure you understand how it works. But ultimately you will find the real value of text analysis by implementing it. The case studies in the book thoroughly cover many interesting and useful applications of text data in a wide variety of business contexts. For a majority of them, we’ve provided the data, projects, and miscellaneous material needed for you to try the discussed methods and approaches yourself. This is the best way to learn and apply those techniques and methods to your business problem. Who knows, you may be able to create an innovative solution for a complex problem bothering your organization.

So, why wait? Embrace the challenge, learn how to use the SAS Text Analytics tools, and start building your own solutions. We look forward to hearing some interesting stories from you, the exceptional soldiers – the SAS users.

Text data is real. It is huge. It is coming. Prepare yourself for the battle!

Coauthors Goutam Chakraborty and Satish Garla also contributed to this post. You can learn more about them and their new book Text Mining and Analysis here.