Top Data Trends for 2018 – Part 1
What data makes possible today is really much more than what data made possible just a few years ago. In this era of digital everything, data now informs and empowers transformative things all around us. This includes data science organizational culture, early warning detection of important phenomena, predictive (forecasting) of outcomes yet to come, prescriptive (optimization) modeling for better outcomes, robotic automation of complex and/or redundant processes, and deep machine learning for enterprise intelligence from rich data assets.
In this first post in a series exclusive to Data Makes Possible, Dr. Kirk Borne, Principal Data Scientist and Executive Advisor for Booz Allen Hamilton, outlines the top 10 trends in Data Science and AI from 2017 that are carrying over into the new year of 2018. In future posts, he will explain the real power of AI and Data Science to augment and assist data-driven decisions and actions in all organizations, regardless of size, industry, or application domain.
1. IoT – The Internet of Context
You could argue that the Internet of Things (IoT) is not a new trend or new thing. You’re right! In its essence, IoT is not new. However, what is new about IoT is the new set of value propositions that it offers. We now see ubiquitous sensors emerging on many different things, from your clothing to orbiting satellites. This “internet of everything”, when outfitted with data analytics and machine learning algorithms at the edge, leads to value-laden concepts like Edge Computing, Edge Analytics, the Analytics of Things, and my favorite: The Internet of Context.
Some have said that the current era of integrated co-existence of sensors, data, and analytics represents a new “Age of Context”. Specifically, streaming data from IoT sensors provide rich and important context to all of our other data sources, processes, and decisions. I would say that the Internet used to be a Thing, but now, Things are the Internet.
2. Hyper-Personalization – True SegOne
Personalization is another well-used and established concept, particularly in marketing and digital user interactions. But, many of those personalized experiences are based on personas (e.g., customer segments). These experiences can now be truly personalized, evolving hierarchically from broad group persona down to SegOne (segment of one).
Hyper-personalization is informed by rich, contextual, intentional, and non-intentional data sources: spatial-temporal data (from mobile/social apps and digital interactions), digital exhaust (from mobile apps and digital interactions), and social trails (from social apps and digital interactions). You get the idea: our apps and digital interactions generate signals that reflect our personal interests, behaviors, preferences, and objectives. Hyper-personalization of services and products attracts our attention and delights us with the right content, at the right time, at the right location, and in the right context. No personas need apply!
3. AI – Augmenting and Amplifying Intelligence
In a way, AI is just another over-used, over-hyped, and worn out buzzword. What makes “AI” new and special is that many (if not most) AI experts are now insisting that we are not talking about “Artificial” intelligence. The latter (which is what we see in movies and hear in media discussions of robots) is more appropriately called AGI: Artificial General Intelligence.
What AI should refer to (and is now referring to) is in fact Augmented, Amplified, and Assisted Intelligence. That’s humans assisting machines (on tasks that require human domain expertise and knowledge) and machines assisting humans (on complex tasks, beyond human cognitive capabilities). That’s the AI that we need and the AI that won’t take our jobs away, but will make our jobs more fulfilling and interesting.
4. Machine Intelligence – Driving Autonomous Things
Machine Intelligence (MI) is essentially the implementation of AI into actionable things: products and processes. MI is seen in automation, chatbots, computer vision applications, and other task-oriented applications. The emergence of MI enables the algorithmic “mathematical” corporation. For those tasks that a machine can do, the machines are now being trained to do them.
MI does not operate within a theoretical cocoon of mathematics. Instead, it is informed and driven by data, from sensors and other sources. Autonomous vehicles are but one example of “self-driving” (autonomous) things that will automate or augment capabilities and drive success within organizations.
5. AR – Placing Data Before Your Eyes
Augmented Reality (AR) is becoming a serious tool in many application domains. These include field operations, manufacturing, disaster response, training programs for complex tasks, search & pick in logistics activities, gamification of learning, and immersive 3D data & information visualizations.
AR is also a major feature of digital twin technologies. Digital twins are changing the game in manufacturing, monitoring and maintaining complex physical objects, processes and systems. When combined with streaming IoT data from a physical asset, the digital replica of the asset enables diagnostic (what happened?), predictive (what will happen?), and prescriptive (how can we optimize what will happen?) analytics to be applied virtually after, before, and during events occurring on the physical asset.
In the next part of Kirk Borne’s Top Trends in Data for 2018, we’ll learn about Behavioral Analytics, Graph Analytics, DataOps, the eXperience Economy and more. Visit Datamakespossible.com on January 8th, 2018 and check out the second half of his list! And, take the conversation to Twitter! Agree or disagree with this list of what to keep an eye on in the new year? Tweet @KirkDBorne using the hashtag #datamakespossible!