Centre for Cognitive Science (COGS)

Prof Takashi Ikegami: The Default Mode Network of the Brain and Transfer Entropy Network


21st Feb 2012

Social network services have been widely recognized after the Egyptian revolution and the Tohoku earthquake in Japan,where the bursting Twitter postings (called tweets) helped the Egyptian revolution and people in Tohoku to get information and connect. But even without major social effects, the Web demonstrates spontaneous bursting behavior. This is caused by the intrinsic fluctuation among tweets as people re-tweet and tweets are consciously orsub-consciously affected by the time lines. We here try to characterize the default mode of the Twitter/Google network to discuss the “autonomy” of the web dynamics as well as the reactive mode which is characterized by the reactive bursts of tweets against the social effects. More specifically, we employ the transfer entropy as an effective measurement tool to characterize the information flow between Google queries and tweets. An analysis on our dataset reveals that when a keyword in tweets become a sink (or source) of entropy flows, it also becomes a sink (or source) in the Google query, respectively. We call the former case, the reactive mode of the Web, and the latter case, the default mode of the Web. Developing the default mode can be a critical step to become a life-like system. Here we hypothesize that the Web can be used as a novel testbed for understanding artificial life in an open environment. References: Takashi Ikegami, Mizuki Oka, Hirotake Abe : Autonomy of the internet: complexity of flow dynamics in a packet switching network, Proc. of ECAL 2011, pp.364-371, 2011. Mizuki Oka, Hirotake Abe, Takashi Ikegami : Autonomous Computation and Flow Dynamics in Web Systems, Proc. of the 2nd International Conference on Morphological Computation (ICMC), pp.76-78, 2011. Mizuki Oka, Takashi Ikegami : The Web as a new framework for understanding the mind, Proc. of the 15th Annual Meeting of the ASSC, Kyoto, 2011.