If you want to evoke a passionate debate these days, just mention that you are working on a “big data” project. It seems as though the mere mention of the term connotes at worst spying or, at best, targeted advertising. But there is a growing chorus within the big data community for the prospects of “big data” being used for social good. There is no denying that Facebook, Google (News - Alert) and even our phone companies collect a lot of information about each of us. The question is: can we leverage big data for the benefit of society and to improve the human condition?
What a decade ago would have seemed impossible is very much a reality today. The pervasiveness of the smartphone and how people are using it has fundamentally changed our ability to leverage real-time communications data to the benefit of our society.
For many people, the smartphone has replaced the computer as the primary device to search for information on the Internet. In addition, when people use a smartphone, it is possible to place them in context of their community and travel patterns. To understand how real-time network traffic and big data (historical usage and traffic pattern) can be leveraged for social good, let’s look at two possible scenarios, each of which can be accomplished without needing to disclose individual user information.
Public Health – Getting Ahead of an Outbreak
Zika is parasite spread by mosquitoes that produces flu-like symptoms and can have grave consequences on a developing fetus, causing microcephaly. To control the mosquito population, local vector control agencies place field traps to capture mosquitoes and then periodically test the mosquitoes captured in the traps. This helps direct public spraying to control the mosquito population. But this approach is slow and reactive. What we have learned from flu epidemics is that there is typically an increase in Google searches of “flu symptoms” that emerge just before or at the same time as an outbreak of influenza. Since Zika is a mosquito-borne pathogen, it will occur outside of times of the normal spread of influenza, but the initial symptoms are very similar to the common flu.
By monitoring mobile searches for any of a number of unique search terms, it would be possible to quickly identify real-time locations where outbreaks may be occurring; thus allowing for a much more targeted response by both vector control as well as public health agencies. In addition, it would be possible to identify the extent to which migration through the area has occurred and to where that population has traveled.
When merged with environmental data like wind patterns, temperature and precipitation, public health agencies can be extremely targeted as to where they deploy resources and the nature of the resources to deploy. But such a targeted and immediate response is only available through the use of real-time network traffic data.
Public Safety – Disaster Response
Recent earthquakes around the globe have demonstrated the potential death and destruction that they can produce. In response to collapsed buildings, first responders rush in looking for survivors; they often act at their own risk because aftershocks could cause additional damage to already weakened buildings. But, the search for life goes on, often with no knowledge as to the number of potential causalities within a building.
To ensure the appropriate allocation of response teams, public safety agencies could leverage mobile network data. When a mobile phone is turned on, it automatically registers to the mobile network. At this point, the operator knows the number of devices originating from a certain area based on the placement of the cell tower and the parameters of that tower.
By comparing the “last known” number of registrants versus historical network usage, the operator could help guide public safety and relief agencies by understanding the number of known mobile phones in an impacted area. If needed, the operator could also assist in the identification of “who” might still be in a damaged structure, should that level of detail be required.
Balancing Privacy and Social Good
As the conversation around using big data for social good evolves, there will be the question of personal privacy. While the types of examples discussed leverage mobile subscriber behavior – be it search terms or location- the purpose is to understand populations or communities, not to identify individual subscribers. With this in mind, it is easy to mask subscriber identifiers while preserving the information about the population.
Ultimately, the goal is to provide a more efficient utilization and allocation of society’s resources as we work to improve the human condition.
About the Author
Mike has over 20 years of experience in the communications industry. He is currently responsible for Service Provider Marketing at NETSCOUT. He began his career at PacBell (now part of AT&T (News - Alert)) where he designed service plans for the business market and where he was responsible for demand analysis and modeling. His career continued with Lucent technologies where he brought to market the first mobile data service technology. At Alloptic (News - Alert), he was responsible for marketing the industry’s first EPON access solution and bringing to market the first RFOG solution. At O3B Networks, Mike headed up marketing, bringing to market the first MEO-based constellation of satellites for serving Internet service to the other 3 billion on the planet. Mike’s work continued at Cisco (News - Alert) where he helped to define MediaNet (Videoscape) and the network technology transformation for cable operators.