Is Cognitive Computing Ready for Prime Time?By Samuel Greengard | Posted 2016-03-17 Email Print
Radical advances in artificial intelligence, along with greater processing power, are pushing cognitive computing and deep learning into the mainstream.
The cognitive system grabs data from an internal database, as well as from external social media, review sites and others to generate Web pages and content. "It mines information and then bubbles up key concepts," explains Phil Koserowski, vice president of interactive marketing for LHW.
However, the system doesn't benefit only potential guests on the front end. It also delivers detailed data that the company can use for analytics.
"One of the benefits of this system is that we can see the types of queries streaming in and rank by percent the level of confidence it has that the results are relevant," Koserowski says. This enables LWH to improve results over time, while also understanding how customers search, what customers want and what hotel changes can help accommodate their interests.
Koserowski says that cognitive tools and analytics will help LHW develop more robust digital personas and will automatically point customers to the types of places, activities, events and cuisines they gravitate toward. It will work in much the same way Amazon and other retail sites customize pages, and it will also enhance marketing campaigns.
The company currently has more than 26,000 images, 35,000 concepts and 351,000 relationships built into the system.
"We are attempting to reduce friction points and make it incredibly easy for people to find what they are looking for," he notes. "If you're someone who always books suites, there's no need to display 30 room types. If you're a person who loves to scuba dive, it makes sense to show relevant resorts and activities."
Cognitive Systems Will Improve Over Time
Over the next few years, cognitive computing and deep learning will continue to leap forward and further change the way humans interact with machines and the way organizations interact with customers. By plugging in large volumes of data and filtering through data relationships in a far more robust way, this technology will deliver growing insights about behavior, buying patterns, financial and scientific models, health care modalities and much more.
Moreover, these cognitive systems will improve over time using machine learning. The one caveat, Capgemini's Cohen says, is that organizations must consider customer and employee privacy and determine how to avoid abuses and breaches of personal information.
Business and IT leaders should keep pace with cognitive computing trends, experiment with the technology and plug it in where it makes sense, he suggests. This may mean embedding cognitive computing capabilities into Websites, mobile apps, and various other tools and systems, including those that tap APIs from partners and business networks.
"Today, plug-and-play cognitive computing solutions—from cyber-security to customer support networks—exist," Cohen says. "Organizational leaders need to focus on the business value and impact that can be realized by applying new capabilities to deliver lower operational costs, increased revenue, competitive advantage, and new and disruptive business models."