Insider threats are an escalating concern for more companies, be they intentional or not. Looking to better detect insecure behavior on the part of employees and contractors, cognitive computing specialist Digital Reasoning has launched a machine learning platform for analyzing human language data at scale.
Synthesys 3.9 uses Big Data analytics to combine disconnected data into synthesized knowledge about customers, employees and high-value targets in order to achieve better results and mitigate increasing risks.
While organizations have invested significantly to manage the surging volume, velocity and variety of data, it has proven difficult to extract real business value from Big Data. The complexity of Big Data environments means it is difficult to normalize and securely connect information across disparate sources — hampering efforts to apply Big Data to critical processes such as conduct risk management in banking as well as threat detection and targeting in the context of national security.
According to a February 2014 report by Forrester Research, ”Cognitive computing brings together previously siloed analytics approaches like natural language processing, text analytics, and predictive analytics inside machine learning solutions that understand human intention and provide answers with transparent confidence...Over the next few years, we will see a growing number of ‘prepackaged’ cognitive solutions — often running in the cloud — to hide the immense complexity of the underlying technology for the end user and make it more economically feasible to experiment.”
Synthesys 3.9 provides capabilities to detect key risk and performance indicators in human communication, and automatically connect and aggregate information based on those indicators.
“We are stoked to announce the availability of Synthesys 3.9 which delivers compelling innovation and critical new capabilities in areas of interoperability, extensibility and security,” said Marten den Haring, senior vice president of products at Digital Reasoning, in a statement. “Our customers are hungry for enterprise-wide big data architectures that empower impactful analysis and discovery, and this release continues to prove why cognitive computing is disrupting traditional enterprise software solutions.”
The platform can monitor data streams in real-time with configurable key indicators, be they tracking online news sources for indicators of social unrest, or monitoring internal employee communication for indicators of fraud and abuse, organizations can define what should generate alerts for behaviors of interest. Key indicators identify multi-dimensional patterns in text by combining learning algorithms and rules engines. Users can review, escalate and explore historical alerts through an intuitive interface with role-based security.
The release also extends Synthesys’ Natural Language Processing (NLP) capability to extract and categorize valuable topics, events and relationships. The NLP framework can easily be extended with third-party entity extractors, data specific preprocessing engines or custom algorithms.
It’s not all machine-based, though. The platform also provides human-curated information about people, places and organizations, so that the system can better organize entities and complex relationships.