Friday, October 14, 2011

Riley Crane, Reality Mining & Red Balloons

Alex Pentland, professor and founder of the Human Dynamics Lab in MIT, is using electronic ID badges, mobile phones and digital media to extract subtle patterns and predict human behavior.  One of his postdoctoral students, Dr. Riley Crane has been assisting him in related areas of this research.

In 2009, Harvard Business Review declared reality mining a "breakthrough idea"and Technology Review stated it to be a "technology poised to change the world."

The Second Channel Of Communication
Research done by Dr. Pentland reveals that there is a "second channel
" of communication between humans that is below the consciousness level.  Research in this area, performed since 2000, seems to demonstrate that "human behavior is much more predictable than is generally thought."
Dr. Alex Pentland
In this same article published in Nature, Vol 457:29, January 2009, Professor Michael Gazzaniga of the University of California, Santa Barbara, states,
A gazillion experiments show that I can flash something at you so fast you don’t see it, yet the information does bias you towards one decision as opposed to another.
According to these professors, the rationalizations as to why we do something come later; they are not the determining factors for that decision.   Of course, not all agree with this view of human behavior.  Bernard Rime of the Catholic University of Louvain in Belgium, views it as too simplistic.  Mark Buchanan in this Nature article, Secret Signals, quotes Rime,
...he says that the relative simplicity of predicting some kinds of human behavior may paint a false picture due to the formidable diversity and contextual variety of human signalling. The link between environmental cues and displayed responses is sometimes simple, as in the call center experiments. But more frequently, Rimé suggests, many other factors interfere, such as past experiences, personal habits, inherited traits, intellectual abilities and shared knowledge about the situation. “In this case,” he says, “predicting people’s behaviour becomes quite a challenging task, if possible at all.”
Riley Crane & Crowdsourcing
Riley Crane is a Society Science Postdoctoral Fellow in Professor Pentland's Human Dynamics Group at the Media Lab in the Massachusetts Institute of Technology.  But whatever view one takes of what the truth of human behavior is, more observation is the key.  How can large groups of people's behaviors be studied?  The obvious answer seems to be social media in all its forms.  In a January 2009 lecture in Hannover Germany Dr. Crane stated,
We're starting to understand global collective trends based on a fundamental understanding of how individuals behave.  So we're looking into statistics of individual human behavior and start to pull out details of how this forms transfers to a global level.
Riley Crane
Some may object that it is not possible to achieve quality assessments on the Internet.   In the "older" days when there were a lot of individual websites that were considered authoritative or important, Google PageRank service worked well.  But now what is happening more and more is that users are not just consuming information, but creating it.  Even further, users are depositing their information in large central repositories like YouTube, Facebook, Google Plus, etc.  Although for example, YouTube videos can still have links to other sites, this is beginning to change the previous structure of page rank.
Google's pagerank structure

Dr. Crane is influenced by Dr. Richard Simon, a Biostatistician and head of the Molecular Statistics and Bioinformatics, at the National Cancer Institute.  Crane quotes Simon,
The information rich world,  the wealth of information is  a dearth of something else, a scarcity of whatever else that information consumes.  What information consumes is not obvious.  It consumes the attention of it recipients.  Hence, a wealth of information, creates a poverty of attention.
This quote naturally begs the question as to how do people choose what to watch on let's say YouTube?  To Crane, it is about studying trends.  One example Crane looks at is Google Trends, especially how the Flu moves through the world in a period of a year.  We include a visualization we obtained from Google's website. If you cannot see the embedded video, here is the link: http://youtu.be/5li8VEYt65I.

The idea was to use search queries in Google to determine how many people had the Flu.  This could be far faster than the traditional methods of determining this information, and even then, it would be two weeks old.  Google compared their queries to the information provided using traditional methods and it was very close.  Of course this information took a day and in the future, it could be done in real time.

But the real interest of Crane is to understand the ebb and flow of daily activities for groups of people.  One of the most interesting things that Dr. Crane found is that mathematical models that are used to map an earthquake, actually also reliably predict how a YouTube model will be received.  Science Daily in 2008 stated when describing Crane's research,
...he discovered that the fading of attention for viral videos can be described with the mathematics used to model aftershocks in earthquakes, so-called “Epidemic Type Aftershock” models. “I find it fascinating that a social system ostensibly works according to particular rules, just like a physical system, and therefore becomes mathematically comprehensible,” says Crane describing his interest in “sociophysics”. He used deterministic power laws to mathematically reproduce the phenomena observed on YouTube. These are scale-independent, meaning that the function’s basic properties also remain unchanged despite changes to the scale. His model can therefore be used singly to recognize developments that could lead to a mass phenomenon using tendencies in the system – in the case of YouTube, an increase in viewers for a particular video. And all this even before the development has been realized by a critical mass of individuals.
Crane divides patterns into two types, endogenous (caused by something from the inside) and exogenous (a pattern caused by an external effect, such as an earthquake or a recommendation by Oprah). He plots them by these shapes,

click to enlarge


Just like one earthquake can trigger subsequent earthquakes so it is in how a YouTube video will spread, a series of cascading events, which spreads until it has gone through a massive part of the entire social system.  In a recent paper, titled, New power law signature of media exposure in human response waiting time distributions, Crane concludes that shocks in systems, whether endogenous or exogenous, produce "...clearly identifiable footprints in time."  This leads him to conclude that "...these findings may open a new frontier for the quantitative study of individual and social activity."

The core of Crane's thesis is stated clearly in a paper he wrote in 2008, entitled, Robust dynamic classes revealed by measuring the response function of a social system.  Crane seeks an underlying universal behavior in social internet systems.  This is the heart of complex systems theory.
A tenant of complex systems theory is that many seemingly disparate and unrelated systems actually share an underlying universal behavior. In the digital age, we now have access to unprecedented stores of data on human activity. This data is usually almost trivial to acquire—in both time and money—when compared with “traditional” measurements. If the complex behavior in social systems is shared by other complex systems, then our approach, which disentangles the individual response from the collective, may provide a useful framework for the study of their dynamics.
Throughout this paper and other lectures, Crane emphasizes the obvious commercial value of being able to predict a bestselling movie or book, a complex-theory-intelligent way to direct a marketing campaign, but it does not stop there. If indeed, all social systems and natural systems follow a universal pattern or patterns, then not only can revolutions and social upheavals be predicted, but also controlled or even prevented from happening.  In this dark side of these algorithms, lie a quest for power.  We are not saying that Crane is looking to aid this kind of freedom-inhibiting behavior, but he, we are sure, is aware of the potential for abuse from these mathematical models.

In our next installment, we will speak of DARPA's red balloons.

No comments: