We are living with increasingly large quantities of data, with at least 1.8 zettabytes worth of data to be created and replicated this year alone. In addition, at least 70 percent of business-relevant information lives in unstructured data.

The advent of multiple platforms for communication and content creation, coupled with the sheer variety of formats and mediums at play, including audio, video, blogs, micro-blogs and social media messages, has exponentially increased the amount of that unstructured data.

Today, businesses are stuck with mountains of data to manage, and are challenged with making sense of the data and deriving value out of it. To stay competitive, businesses can no longer afford to ignore the critical role that unstructured data plays in driving their day-to-day decision-making processes.

Here is where text analytics and text mining tools play a key role. Text analytics enable businesses to maximize the value of information within large quantities of text that is generated, acquired or exists in repositories by extracting relevant information, interpreting it, then structuring information to reveal patterns, sentiments and relationships among documents.

For example, unstructured transcripts, notes and documents describing business activity can provide important insights about customers’ habits, tastes, product use and support requirements, employee work habits and performance, as well as business process efficiencies and failures.

In most business scenarios, text mining can have an immediate and measurable effect on CRM policies. Information extracted from various customer touch points such as email, transactions, call center communications and customer surveys can be used to predict customer churn as well as help in building customer up-sell and cross-sell models.

Beyond refining internal business processes, the usefulness of text mining on gaining market insights cannot be understated. Daily analyst reports, news articles and company press releases contain vital information about market shifts and changes in the investment environment, upon which trend analyses can be performed utilizing systematic approaches through text mining. The same results can be gleaned from qualitative data found in blogs, commentaries, open-ended survey questions, focus groups and interviews.

Applications with greater socioeconomic implications

Text mining can be combined with other data mining tools to look for patterns and anomalies in huge amounts of text data, bringing about great benefits to critical industries such as health care, insurance and even the public sector. It can be performed for text extraction on patent reports, journal articles, and other medical publications, reducing the amount of time needed for extraneous and redundant research. Association models can be used to detect patterns such as in the testing of drug symptoms, therefore minimizing time spent in the drug discovery process. Furthermore, text mining’s capability to sift through massive amounts of data and emerge with verifiable patterns and associations represents a tremendous opportunity for national security agencies to get ahead of criminal and terrorist organizations.

At the end of the day, imagine what your business could do if you could harness the insights hidden beneath that 70 percent of unstructured data. Picture how it would improve knowledge sharing and facilitate decision making if useful content was easy to find and use, and automatically included in analytical processes. In a world containing more and more data, text analytics will only become more valuable.

By David Hardoon, Principal Analyst, SAS Singapore