Category: Client Forum

First be a Scientist – A Big Data Approach


Big Data CitiesJameson Toole is a Data Scientist and PhD candidate at MIT. We were thrilled when he agreed to speak at the Client Forum in Boston as Boston Financial celebrates our 40th Anniversary.  Our company and industry have changed in countless ways over the past 40 years, and Jameson’s talk underscored this in covering a hot topic that didn’t exist just a mere five years ago.

Jameson is working on a fascinating project as he studies big data and cities, and the systems that sustain people.  He says that 50% of people now live in cities and that number is growing.  This is a great time to study this, as everyone is carrying a cell phone around, and the aggregate data that can be analyzed is incredible.

Jameson said that a data scientist must be knowledgeable in many things, and have contextual awareness.  A data scientist should not be isolated to crunch numbers.  They need to be part of strategic meetings.  At the end of the day the infrastructure is not very important.  Success is driven primarily on context and in essence, asking the right questions of the data.  If you have context, and understand the need, you can form good questions to answer with the data.  His key message is to be a scientist first.

Jameson described a number of fascinating case studies and analysis he has been working on.  One illustrative example shows a visualization and time lapse animation of all the calls in Boston day or night, over a period of months.  He paused the animation at an obvious spike in calls, and pointed out that happened on Valentine’s Day evening, and theorized on what the calls might have been about (e.g. dinner reservations).

In another example, he referenced an app installed on his phone.  It records his location every 15 minutes, and then he can analyze his movements around the city over time.  He says that his team analyzed data from a large population of people and determined people have 4 primary locations on an average day, perhaps things like home, work, gym, or a coffee shop.  And while everyone has a different set of 4 locations, they visit them in the same order.  In expanding to the top 6, there are over a million different ways to move between 6 locations. There are over 1 million ways for individuals to travel between 6 locations, yet 90% choose one of only 17 patterns.  Understanding the behavioral patterns of movement of 90% of the people could lead to some truly innovative, breakthrough applications and solutions.

At the end of the day, Jameson says that people are very predictable and there is an amazing set of data that can be analyzed to do incredible things.  There are business applications that will allow location services, directed advertising, and the like.  In our industry, it could be about custom selling methods, location services, and predictive analytics on consumer sentiment.

But he wants to focus on people and cities, solving problems (e.g. traffic congestion) and improving life experiences.  I think that is a great application worthy of praise.  Perhaps even a few bucks if his projects show up on KickStarter someday.

Remember his name.  Jameson Toole.  A scientist, first.

Michael Nadeau

Michael Nadeau

Mike has been in Information Technology and Financial Services for over 20 years. He is currently responsible for a number of teams in IT, such as Project Management, Software Development, Business/Systems Analysis, Database Management, Product Support, and Knowledge and Collaboration Solutions.

Related Articles

Looking at the Big Picture: Leveraging the power of data »

By Arthur Dunn | June 8, 2015

Cybersecurity, Intermediary Oversight Hold Their Ground for 2015 »

By Mike McNeill | December 16, 2014

Revisiting the 2013 Client Forum »

By Kristin Ferguson | August 7, 2014

Your comments mean a lot to us. We want to hear your perspectives, but please know that this section is being moderated and we reserve the right to edit or delete content at our discretion. Please keep your comments respectful and relevant.

+ Post a comment

Your email address will not be published. Required fields are marked *