Utilizing Cutting-Edge Unstructured Data AnalyticsBy Guest Author | Posted 2015-03-05 Email Print
Re-Thinking HR: What Every CIO Needs to Know About Tomorrow's Workforce
A major challenge is finding ways to harness unstructured data—such as call center logs, blogs and social media posts—in order to overcome business challenges.
By Durjoy Patranabish and Shreya Sharma
As technological advances provide more efficient ways to connect and personalize products and services, the expectations and demands of customers are increasing. They want ever-more personalized products with ever-faster turnaround times.
Companies now need to find new ways to make smarter and faster decisions, and obtain superior insights into customer needs, behaviors and profitability. To remain competitive, agile and innovative, organizations must make analytics an integral part of their strategic planning, investing time and talent to explore the best ways to utilize data. Among the biggest challenges is finding ways to harness unstructured data—such as call center logs, blogs and social media posts—in order to overcome modern business challenges and increase revenues.
At a macro level, unstructured data encompasses any data that cannot be contained in a database or fit within a predefined data structure. Until recent years, many didn’t think it was possible to fully utilize the unstructured aspect of big data. They didn’t think analytics solutions existed to extract meaningful data from unstructured text, audio and video, nor did most companies have the infrastructure or capabilities to experiment with unstructured data.
Today’s market is very different. The surge of digital and social media data enables companies to glean a 360-degree view of their customers, and the ability to utilize unstructured data in a reasonable and realistic manner is within reach.
On average, 70 to 80 percent of an organization’s data consists of unstructured data. If analyzed correctly, this volume of data can contribute about 20 to 30 percent of total insights, enabling companies to understand customer mindsets, purchase preferences and sentiments far better than they could have done by using only structured data.
While there is no shortage of unstructured data, the challenge comes in accessing that data and transforming it into something usable and actionable. The following technological advances are helping companies converge structured and unstructured data to overcome business barriers:
Text mining and neuro-linguistic programming (NLP) help companies tap unstructured data to analyze customer feedback, such as emails, blog posts, customer comments and queries, so they can extract meaningful insights. For instance, these capabilities have the power to:
· Manage real-time data feeds
· Convert unstructured text into structured data, telling companies "who" is being discussed, "what" is the context of the conversation, and whether that feedback is positive or negative
· Reduce manual efforts and turnaround time significantly
· Scale solutions to continually manage voluminous data
Artificial intelligence (AI) has been used with research methodologies in medicine, robot control, defense and the remote sensing industries. One key distinction is that AI diagnostic methods emphasize the algorithm—as opposed to standard predictive analytics techniques—and are able to predict business scenarios and trends by leveraging historic data.
Machine learning enables rapid processing of large amounts of customer-centric data, including customer conversations in the form of calls, email and chats. It has the ability to handle huge volumes of data, enabling companies to incorporate multiple data streams without compromising the accuracy or relevance of the results. For instance, companies are using machine learning to build a single view of customers so they can develop the next best offer or predict churn.
Sentiment analytics allows analysts to tap into the world of social media and combine it with cutting-edge algorithms such as NLP in order to decode complex human interactions and evolving languages to understand what customers are really saying. For instance, movie makers can use algorithms and computational models to learn about their audiences so they can script a movie that will have maximum impact.
High-performance computing has the ability to handle huge volumes of data on an almost real-time basis, thereby transforming the way data used to be analyzed. As unstructured data consists of nearly 80 percent of a company’s total data, the boundaries between unstructured data analytics and high-speed computing powers continue to blur. The large volume of unstructured data combined with its ever-increasing velocity demands that companies move away from traditional computing systems. Further, as algorithms become more complex in nature, executing them requires higher computing power to be effective.
What’s Next for Unstructured Data Analytics?
Recognizing the critical benefits of unstructured data, companies are ready to invest in research and development to solve complex challenges. The opportunity cost of ignoring unstructured data is now too high in this globally competitive environment.
Technology advances in unstructured data analytics are helping companies across industries understand customers, making sales agents more efficient and enabling maximum returns from marketing activities.
Until recently, statistics and data management have served as two prominent pillars of analytics. But now, with increased computing power and improvised techniques, unstructured data is quickly emerging as the third pillar.
We’ve seen only the tip of the unstructured data iceberg. Innovative companies with the foresight and desire to dig deeper have the potential to change the way we live.
Durjoy Patranabish is senior vice president, Analytics, and Shreya Sharma is consultant, Analytics, at Blueocean Market Intelligence. The company is a global analytics and insights provider that helps corporations obtain a 360-degree view of their customers through data integration and a multidisciplinary approach.