(eToro Blog) We like to think that we are unpredictable, unforeseen, and masters of our own destinies. In fact, nothing could be farther from the truth. In the “Reality Mining” research (conducted in 2005 by Dr. Nathan Eagle under the supervision of Prof. Alex “Sandy” Pentland, in the Human Dynamics group of the MIT Media Lab), a large scale quantity of data regarding the behaviors and activities of residents of the Massachusetts Institute of Technology and Harvard University area was recorded. Surprisingly, researchers were able to show that the rules that govern the dynamic behaviors of the participants were not as chaotic and complicated as thought initially. In fact, this research has demonstrated that future behaviors can be predicted with high accuracy, based on the patterns of the past activities.
The following chart is taken from a publication by Eagle and Pentland that was published by the American National Academy of Science, and which shows the repetitive patterns of communication between the participants. With this information, future interactions can be predicted, with high success probability:
This research gave way to a multitude of additional projects, all focused on the goal of developing methods for analyzing and predicting human behaviors in a variety of fields. What can we learn from such results? And why are they so interesting? The answer is simple as it is potent: We want to get a head start.
Knowing in advance the likelihood that people walking on a given street would be hungry can be very helpful for an entrepreneur who is looking for a new place to open his restaurant. Similarly, knowing the patterns of internet usage through the hours of a day can assist an ISP to design its bandwidth allocation, or the mobility patterns of the population would be very helpful to a mobile carrier in considering where to deploy a mobile cell tower.
In trading and finance, knowing when people are more likely to perform trading transactions can serve as invaluable information, as it provides important knowledge concerning the expected dynamic liquidity of the markets. In addition, having this information segmented across various types of traders can help a broker or a hedge fund decide when to put its desired orders to the market. Based on this premise, traders armed with this information, can predict the likely future behavior of the hedge fund, and try to gain leverage from the information.
These, perhaps, were also the thoughts of Dr. Eagle; several years after completing this groundbreaking research, he established his own company “txteagle,” which collaborates with over 200 mobile carriers in more than 100 countries, and generates valuable commercial insights from behavioral data.
Copyright 2011 eToro Blog
