It’s time to join the AI revolution

There’s a reason artificial intelligence (AI) is lauded as a technological shift as revolutionary as the smartphone. Computers can analyze vast amounts of data more quickly and more accurately than humans—and even help us predict consumer behavior.

We all know online retailers are able to use our digital footprints to predict what we might be in the market to buy and then serve up personalized ads and promotions. You can put this same intelligence to work for you.

If you’re trying to make sense of and identify patterns in large volumes of data, it’s time to learn about machine learning. AI can help you contain the costs of big data projects, leverage the value of data that was previously collected manually, collect real-world intelligence about your customers, reduce the lag time in reporting findings—and allow you to take informed action quickly.

 

A primer on AI, machine learning and deep learning

Curious but confused? Here’s a quick overview to help you understand some of the terminology.

• Artificial intelligence (AI) is intelligence demonstrated by machines. Most often is it used to describe computers that can mimic cognitive functions—such as learning and problem solving—that we normally associate with the human mind.
• Machine learning is a branch of AI that uses computer algorithms and statistical models to make predictions. Instead of using explicit instructions about how to perform a specific task, machine learning relies on the recognition of patterns in sample or “training” data to make inferences and predictions. Thus, machine learning is also referred to as predictive analytics. A widely used application of machine learning is email filtering—which uses recognized data and patterns to predict which messages shouldn’t pass your spam filter and land in your inbox.
• Deep learning is a subset of machine learning that uses multiple layers of data to replicate the hierarchical neural networks and processing ability of the human brain to recognize patterns and derive meaning. Deep learning has been applied to fields such as computer vision, speech recognition, audio recognition, language processing, and social network filtering.

 

Fewer barriers to entry

While AI, machine learning and deep learning have been around since the 1950’s—if only theoretically—three recent developments have made them more accessible: a method which allows software neurons to teach themselves by layering learning, the repurposing of visual processing capabilities originally created for video games, and the sheer volume of data now available—because computer brains feed on large volumes of data.

It’s getting easier and easier to join the cognitive technology revolution. But it can be hard for newcomers to figure out how to get started. Perhaps the easiest way is to think of such predictive analytics as traditional analytics on steroids. For example, if performing regression analysis—for example, how “likelihood to recommend” is impacted by changes in wait time, price, quantity purchased—you can automatically run every possible combination of predictive variables as AI allows for quick and easy processing of huge volumes of data.

 

Examples of how to get started

With the exponential rise of personalized messaging, fragmentation of media channels, and rapid growth of social media, the need for timely market intelligence and compliance monitoring is growing exponentially. And this is a great way to get started with AI.

At EurekaFacts, a Washington, DC area market research firm which serves both government and the private sector, we aim to provide you with actionable consumer insight to drive social impact. For example, we’ve developed a market intelligence and consumer product compliance monitoring system called EurekaFacts PinPoint that collects, stores and analyzes real-time intelligence of entire industries. Our first application monitored tobacco product advertisements across print, online and social media platforms within the U.S. PinPoint provided ongoing monitoring and identification of ads, labels, and packages; coded potential violations of government regulation limiting tobacco advertising; and shared the results with those who monitor regulatory compliance.

Likewise, to assist our clients in understanding consumer sentiment, we’ve developed an AI tool called EurekaFacts Sentiment Analyzer which allows us to reach out to a social media platform such as Twitter, collect available interactions such as the most recent set of fifty tweets, then examine the content for tone and geographic location of origin, and produce a map of positive, negative and neutral comments. When set in a time-based visualization, reactions to different topics can be detected as well as their changes over time.

 

 

Article adapted from the Harvard Business Review Overview of Machine Learning and AI.