Complexity Analytics and Public Policy : Cautions and Opportunities Going Forward
these data , though , because this would not be an exercise in statistics . Some Big Data databases are simply indiscriminate amalgams of huge amounts of data from wherever . In looking for bifurcations , the data must be FROM the systems being analyzed , so matching data to source is another bit of retrofitting that would be required .
What I am suggesting is that searching existing databases and social systems research archives for bifurcation occurrences could lead to new and exciting discoveries about the phase space of social systems . It could also uncover volumes about orthogonality and other aspects of the topology of social space and help unravel the mysteries of attractors giving us access to points of stable and unstable equilibrium in the phase space of complex adaptive social systems with all the associated predictive value .
Modeling Complex Adaptive Social Systems
Genetic modeling is a state spacebased tool of complexity analytics that is very satisfying to the purist as it requires doing very little damage to the assumptions of complexity science . Modeling complex systems is a conundrum , however , especially in the case of complex social systems analysis , in that the researcher is creating the dynamics being observed and attributing to the results some meaning for real-world complex systems ( Miller & Page , 2007 ). This is particularly problematic for fitness landscapes because such landscapes cannot help but be arbitrarily carved out of segments of state space with little or no discipline imposed by an understanding of the underlying topology . In fact , the topology is created by the researcher ’ s a priori understanding of the nature of the system . Complexity science treats the system and its environment as one interactive phenomenon from a dynamic state space perspective . To correctly integrate system components , we need to understand the topologies of the spaces the various system components occupy .
The topologies of physical spaces are well understood ; however , our awareness and understanding of the topology of social space is in its embryonic stage at best . Best efforts to understand the topology of social spaces have focused on networks and have associated the topology of social spaces with network topology . That direction is interesting and promising . This is less problematic for agent-based models ( Axtell , 2000 ; Batty , 2005 ; Reggiani & Nijkamp , 2009 ). While it is true that agents , environments , and rules of interaction are selected , designed , and defined by researchers , the map between these elements and the real world is much more apparent . Conceptually , however , the grids on which agents move in agent-based models are usually mathematically defined distributions that do not begin to capture the deep structure of real social space . Underlying social networks , on the other hand , are maps of reality and they provide powerful insights into the diverse topologies of social space ( Epstein & Axtell , 1996 ).
5