Unprecedented proliferation of applications creates extraordinary opportunities for mining vast amounts of data about consumer behavior patterns
Unprecedented Proliferation of Applications Creates Extraordinary Opportunities for Data Mining
In 2008, when Kaveh Pahlavan, professor of electrical and computer engineering at WPI and director of the Institute's Center for Wireless Information Networks Studies (CWINS), convened 50 invited experts from industry and academia for the first International Workshop on Opportunistic Radio Frequency (RF) Localization for Next Generation Wireless Devices, only a handful of applications were available for those devices. Apple had introduced the iPhone the previous January, but its App Store had yet to debut.
Two years later, there are over 100,000 applications for the iPhone and more are being added all the time. About 10 percent of those apps employ localization technology, and its rapid development has generated extraordinary new opportunities for mining and utilization of data about consumer behavior patterns that will drive the industry's technological innovation for the foreseeable future.
That was a central conclusion when Pahlavan recently hosted the second international workshop. The 50 invited experts spent a full day discussing how swiftly localization technology has evolved and what that rapid evolution implies for the future of the industry.
"We are living in a world flooded with information," Pahlavan said. "When we held the first conference the pressing issue we focused on was availability. But in the past two years technology has made extraordinary strides, and it is continually progressing. Coverage is constantly increasing and so is accuracy, while applications are growing exponentially.
"I firmly believe location and presence will be cornerstones of most successful wireless applications for the foreseeable future," said Richard Lynch, executive vice president and chief technology officer of Verizon, who was the first communications leader to adopt code division multiple access (CDMA) as a wireless access technology in 1998. Noting that long-term evolution (LTE), the latest standard in mobile network technology, will become a reality in the United States this year, he observed that it will soon cover a third of the U.S. population.
"The market continues to grow at a rapid pace," Lynch said. "A quarter of all U.S. wireless phone users have used a location-based service. Smart phones currently make up about 20 percent of the wireless phone market, but sales are growing at an impressive 55 to 60 percent worldwide. With this kind of growth, the tipping point for mass acceptance has arrived."
By 2014, he noted, more than 1.5 billion people worldwide will be using local search, a technology that continues to grow at light speed. "Clearly, wireless will be the preferred and primary way in which most consumers and businesses communicate," he said.
The growth of the marketplace opens up a host of avenues for unprecedented use of the technology to the advantage of consumers and industry. Ted Morgan, CEO of Skyhook Wireless, noted that with the number of location-related apps that use location projected to jump from 10,000 to 50,000 this year, new data mining capacity suggests a world of never before available information that is potentially valuable to all kinds of users.
For instance, by monitoring 'where am I' data for a big city, traffic patterns can be tracked and analyzed in myriad ways. To illustrate the point, Morgan noted the capacity to track how certain demographic clusters move. Skyhook has analyzed the patterns of both mobile and pedestrian traffic at the Boston Marathon and at Major League baseball parks, using signals to track how people get to events, where they park, what they do during and after events, and how they return home.
"Think about what applications can take advantage of those patterns," Morgan said. "It's information that is useful to transportation planners and to advertisers."
Not only can Skyhook track the movements of users, but the company also analyzes locations as precisely as 25 meter-by-25 meter "tiles" in any location, noted Farshid Alizadeh, chief scientist at Skyhook. "We can rank a tile within a city, a city in the world, or a tile in the world," said Alizadeh. "We can track trends and rank what is getting hotter or colder."
The data are reliable, said Morgan, and highly predictive. If a demographic cluster performs in a certain way on a routine Tuesday one week, it is likely to perform the same way on Tuesday a week later. "No one has ever been able to track this many users with this level of precision before," he said.
Marta Gonzalez, assistant professor of civil and environmental engineering at MIT, and Guanling Chen, assistant professor of computer science at University of Massachusetts Lowell, offered complementary presentations about the use of wireless technology to analyze human movement. Gonzalez' research uses statistical physics to integrate and analyze information generated by human dynamics. She demonstrated how transportation patterns, derived from wireless signals, can be layered upon data about commercial, recreational, and professional traffic patterns, offering insight into human movement in urban landscapes. Chen's research analyzes the behavior of users in social networks. He described the use of metrics for what he called "friendship modeling," the analysis of the shared interests and proximity of subscribers to social networking services.
Hannu Hauppinen, director of business development for the Nokia Research Center, talked about three promising prototypes developed by Nokia, the world's largest manufacturer of mobile phones. A mobile social networking program called Nokia Find & Connect links members of a social network within a company to each other and physical resources, addressing common time-consuming workplace problems such as finding meeting rooms, resources and each other. Nokia's High Accuracy Indoor Positioning (HAIP), which has potential applicability in retail, transportation, and sports arena settings, uses existing WiFi networks, directional beacons, and a positioning server to track individuals and objects. And Nokia's Instant Community creates location and context-based networks between people automatically, with no user actions required.
Workshop participants also discussed localization requirements for homeland security and emergency responders; melding air interfaces, bands, planes and technologies; developments in satellite navigation and wireless spectrum; the role of standardization; statistical modeling and the capacity to predict human behavior; and many other topics.
This invitational workshop was sponsored by Verizon Wireless, Skyhook Wireless, and Via-Telecomm. Videos of the presentations and additional details about the workshop are available at the workshop website.
Kaveh Pahlavan, professor of electrical and computer engineering at WPI, established the Center for Wireless Information Network Studies (CWINS) in 1985. It was the first research program in modern wireless networks in the United States and for the past 25 years it has played a unique role in the growth of the wireless LAN industry by publishing pioneering textbooks, seminal visionary papers and patents, and an international journal, as well as organizing key workshops and conferences.
August 18, 2010
Contact: Michael Dorsey, Director of Research Communications, +1-508-831-5609, email@example.com