In a budget-constrained federal environment, a common mantra is “do more with less.” With Digital Reasoning’s machine-learning platform, Synthesys®, it is now possible to augment analysts by enabling these valuable individuals to quickly analyze troves of data that would normally have taken many thousands of analysts an inordinate amount of time to complete.
Analysts today are often still unable to efficiently mine their poorly structured data (e.g. social media, email communications, instant messaging, etc.) This presents a real analysis challenge for policy and decision makers, as information isn’t transformed to knowledge quickly enough, and when expedited, the information isn’t always as accurate as expected.
For example, Twitter sees about 500 million tweets each day and has more than 230 million accounts. 125,000 new accounts are created on Tumblr every day. And the world creates 2.5 quintillion bytes of data per day from data sources like sensors, social media posts, and digital photos. This is information overload for any analyst and quickly leads to both inaccurate decisions and loss of fidelity within the data itself. Having a lot of data available simply doesn’t translate to valuable and actionable information, let alone real knowledge.
But, at the end of the day, it is people who have to make the decision, take action, interdict and mitigate the risk. Acknowledging this reality, and understanding the opportunity to apply new technological approaches to analytics and analysis, Digital Reasoning has amassed over ten years of real-world experience backed by an industry leading research and develop team looking at ways to empower the analyst to think through larger amounts of data, make better decisions, quickly, and with greater accuracy – allowing people and machine to work together and as a result, become more intelligent. That’s the design philosophy behind our machine learning analytics platform, Synthesys.
In a book titled, “Thinking, Fast and Slow”, world-renowned psychologist Daniel Kahneman describes two systems of thinking – system one and system two. System one is fast and automatic and system two is slow and effortful.
We can think of Synthesys as system one; it simulates human reading and comprehension of contextual information via patented, machine-learning algorithms. It takes poorly structured information and structures it into a useable knowledge base. As an example, Synthesys can quickly read and analyze 200 million documents and show how all of the people, places and things are related to everything else. It is able to isolate the signal from the noise by identifying all the relevant and valuable events from each file and correlating that information with every other file being analyzed. This creates a level of situational awareness that has historically been difficult to construct as that same effort would require thousands of analysts working for years to complete. Synthesys never sleeps. It reads everything, it forgets nothing and gets smarter the more an analyst uses it.
Analysts represent system two; they can take the resulting knowledge from system one and do the very necessary, effortful things like draw conclusions, inferences, and ultimately make the required decisions. With the facts in front of them, analysts can now spend less time reading and more time collaborating and sharing the resulting knowledge.
Analysts can now ask questions that would have been impossible with legacy approaches like pattern recognition or rules engines. And, since our products are built on open source software with open APIs, the life cycle operations and maintenance costs fit within constrained IT budgets.
Both Synthesys and the analysts contribute something unique and necessary to efficiently answer difficult questions. And, unlike other systems that read and analyze information and stop there, Synthesys goes further. It reads, resolves and reasons, and then gets smarter as the analyst uses the information and makes decisions – which creates an infinite loop of learning between the analysis and analyst.
Combined, both the analyst and Synthesys represent a more intelligent way of thinking, analyzing information, and transforming data into knowledge, which allows us to do more with less.
To learn more about Digital Reasoning and our platform Synthesys, visit us at: