First published 2 July 2012
Characterizing Autonomy in the Web via Transfer Entropy Network
Mizuki Oka, Takashi Ikegami
Using the idea of transfer entropy (TE), we study autonomy and information flow on the Web and the newly defined TE network. The Web shows rich and complex autonomous network dynamics. Social network services (e.g., Twitter or Facebook) are now becoming a major source of Web dynamics in addition to the Web search services (e.g., Google). It is widely accepted that Twitter messages (called "tweets") and Google search queries react strongly to significant social movements and accidents, which are often characterized by bursting patterns in the time sequences. We call this the reactive mode of the Web. On the other hand, the Web dynamics, without the significant social events, seem to have an intrinsic rich dynamics, which we call the default mode of the Web. In this paper, we study the default mode of the Web system, which we characterize via a TE network. The amount of information flow transferred between different sequences of Google queries as well as Twitter keyword frequencies is investigated and we compute a TE network among Twitter keywords. We then discuss that the default mode of the Web can be characterized by the "breathing" dynamics of the TE network over a scale of a few weeks. We further use this idea of the default mode to install autonomy into generic artificial life systems.