Mattersight Issued Patent for Machine Learning-Based Trend Identification

Mattersight Corporation, the pioneer in personality-based software applications, was recently issued its 39th patent which uncovers and delivers trending topic identification from contact center interactions. The patent, Trend identification and behavioral analytics systems and methods, is the first issued from the trending topics product family and provides a machine learning-based framework for analyzing data trends.

Traditionally, efforts to track and identify emerging topics include manual processes such as keyword-based searches which require an individual to know what they are looking for. With the power of trending topics lying in the ability to identify and respond to them in a timely manner, these methods are inefficient and can be less effective.

Using machine learning algorithms, this new patent allows for the automated identification of trending conversational topics in real time. Scanning a stream of numerous interactions and data types – phone, social media, etc. – trends are identified based on their frequency and co-occurrence without the need to reference a library of pre-defined terms. These trending terms are then displayed to a user along with rich data that provides greater detail into the terms and can generate alerts.

“We’re really excited about this patent, which is the first of its kind for us,” said Mattersight CTO Chris Danson, one of the solution’s inventors. “We’ve found that with many of our customers, there was a need for deeper insight and the ability to go beyond basic trend identification by leveraging the power of machine learning. We’ve seen a lot of excitement around this patent and its ability to self-discover conversational topics that customers are talking about and how these topics impact business outcomes.”

With greater access to enormous and often complex sets of data, there has been an increasing interest from organizations to be able to identify and track emerging trends or topics that could otherwise be overlooked. For example, one of our customers discovered that immediately after deployment there were password conversation topics related to password issues that were largely associated with their customers not making a purchase. The application quantified these issues both in terms of scope and their effect on revenue, providing them with information to resolve the issue with their technical stakeholders.

Leading Fortune 500 organizations use Mattersight’s suite of personality-based software applications to improve customer experience in their contact centers. Using a NASA developed personality model and speech recognition algorithms that mine data from over one billion customer service calls, Mattersight is able to analyze a customer’s personality and behavior traits in seconds and pair them with an agent they are most likely to have chemistry with.

Looking ahead, Mattersight’s data teams are focusing on how to leverage trending topics in predictive modeling while they continue to evolve the current descriptive analytic capabilities.