Abstract
Recent developments in sensing technologies have enabled us to examine the nature of human social behavior in greater detail. By applying an information-theoretic method to the spatiotemporal data of cell-phone locations, [C. Song et al., Science 327, 1018 (2010)] found that human mobility patterns are remarkably predictable. Inspired by their work, we address a similar predictability question in a different kind of human social activity: conversation events. The predictability in the sequence of one’s conversation partners is defined as the degree to which one’s next conversation partner can be predicted given the current partner. We quantify this predictability by using the mutual information. We examine the predictability of conversation events for each individual using the longitudinal data of face-to-face interactions collected from two company offices in Japan. Each subject wears a name tag equipped with an infrared sensor node, and conversation events are marked when signals are exchanged between sensor nodes in close proximity. We find that the conversation events are predictable to a certain extent; knowing the current partner decreases the uncertainty about the next partner by 28.4% on average. Much of the predictability is explained by long-tailed distributions of interevent intervals. However, a predictability also exists in the data, apart from the contribution of their long-tailed nature. In addition, an individual’s predictability is correlated with the position of the individual in the static social network derived from the data. Individuals confined in a community—in the sense of an abundance of surrounding triangles—tend to have low predictability, and those bridging different communities tend to have high predictability.
12 More- Received 5 May 2011
DOI:https://doi.org/10.1103/PhysRevX.1.011008
This article is available under the terms of the Creative Commons Attribution 3.0 License. Further distribution of this work must maintain attribution to the author(s) and the published article’s title, journal citation, and DOI.
Published by the American Physical Society
Popular Summary
Finding and understanding patterns and dynamics of human interactions used to be the research of social science. Recently, however, this topical area has become increasingly multidisciplinary, drawing, not in the least, the interest and contributions of many physicists. Data collected on mobile phone usage or movement of marked currency bills have been analyzed by statistical physicists to reveal patterns of human mobility and their remarkable predictability, which in turn are used to understand phenomena of societal impact such as epidemic spreading. To what extent are the interactions among individuals random or predictable, and does the predictability depend on the context of the interactions, and how? With novel data collected with short-range electronic sensing technologies from hundreds of individuals over a period of several months and a comprehensive analysis of the data, we find some (quantitative) answers to these questions in the context of human conversation events taking place in the office environment of two Japanese companies.
Each individual participating in our data collection wears a name tag equipped with an infrared sensor node, and conversation events are marked when signals that are exchanged between sensor nodes (i.e., the individuals wearing them) are in a well-defined range of physical proximity. The predictability in the sequence of an individual’s conversation partners is defined as the degree to which his/her next conversation partner can be predicted if the individual’s current partner is known. We examine the data and extract that predictability. We find that the knowledge of the current partner decreases the uncertainty in knowing about the next partner by almost a third. Much of the predictability can be explained by long-tailed distributions of interevent intervals—generic patterns seen often in human actions. More interesting perhaps, individuals confined in a community or “group” tend to be less predictable than those bridging different communities.
Many more and interesting questions can be asked following our study. We hope that our unique data and our analysis of the data will serve as a new stimulus to the emerging and active area of “observational social physics.”