Theoretical study on the spreading phenomena in human society

Abstract

Statistical physics is a part of physics that uses mathematical theory of probability and statistics to account large number of populations to solve physical problems. Many laws of modern physical sciences are based on statistical properties, thus statistical physics has provided rich resources to elucidating fundamentals of nature. Physicists who are fascinated by success of physical science extent their interests to more complex one, such as economics, computer science, biology, medicine. Physicists have developed a statistical physical modeling of such problems, which is far-distant from their traditional realm. Social behaviors defined as collective behaviors of group of humankind, distinguished from action of an individual. Social behaviors defined as collective behaviors of group of humankind, distinguished from action of an individual. Interestingly, social behavior shows surprising universalities and regularities regardless of individual's personality. In this context, A term "Social Physics" offers a fresh avenue of understanding such collective behavior. A spreading is one major example of such social physics that refers delivery "things" from a member of society of another. These "things" include both non-physical concept (idea, innovation, knowledge, rumor) and physical one, e.g., infectious disease. This thesis discusses some examples of such spreading phenomena with combination of mathematical model and data science.

Chapter 2 discusses about how science and technology is invented, spread, and generalized via printed publications. The quest for historically impactful science and technology provides invaluable insight into the innovation dynamics of human society, yet many studies are limited to qualitative and small-scale approaches. Here, we investigate scientific evolution through systematic analysis of a massive corpus of digitized English texts between 1800 and 2008. Our analysis reveals great predictability for long-prevailing scientific concepts based on the levels of their prior usage. Interestingly, once a threshold of early adoption rates is passed even slightly, scientific concepts can exhibit sudden leaps in their eventual lifetimes. We developed a mechanistic model to account for such results, indicating that slowly-but-commonly adopted science and technology surprisingly tend to have higher innate strength than fast-and-commonly adopted ones. The model prediction for disciplines other than science was also well verified. Our approach sheds light on unbiased and quantitative analysis of scientific evolution in society, and may provide a useful basis for policy-making.

Chapter 3 discusses about how collective intelligence formulated via Internet-based collaborating environment. Wikipedia is a free Internet encyclopedia with an enormous amount of content. This encyclopedia is written by volunteers with various backgrounds in a collective fashion

anyone can access and edit most of the articles. This open-editing nature may give us prejudice that Wikipedia is an unstable and unreliable source

yet many studies suggest that Wikipedia is even more accurate and self-consistent than traditional encyclopedias. Scholars have attempted to understand such extraordinary credibility, but usually used the number of edits as the unit of time, without consideration of real time. In this work, we probe the formation of such collective intelligence through a systematic analysis using the entire history of 34 534 110 English Wikipedia articles, between 2001 and 2014. From this massive data set, we observe the universality of both timewise and lengthwise editing scales, which suggests that it is essential to consider the real-time dynamics. By considering real time, we find the existence of distinct growth patterns that are unobserved by utilizing the number of edits as the unit of time. To account for these results, we present a mechanistic model that adopts the article editing dynamics based on both editor-editor and editor-article interactions. The model successfully generates the key properties of real Wikipedia articles such as distinct types of articles for the editing patterns characterized by the interrelationship between the numbers of edits and editors, and the article size. In addition, the model indicates that infrequently referred articles tend to grow faster than frequently referred ones, and articles attracting a high motivation to edit counterintuitively reduce the number of participants. We suggest that this decay of participants eventually brings inequality among the editors, which will become more severe with time.

In chapter 4, we perform the in-depth analysis on the inequality in Wikipedia. Advancing from the chapter 3, we perform an in-depth analysis about entire 863 Wikimedia projects including every language edition of Wikipedia, Wiktionary, Wikisource, Wikivoyage, and so on. This comprehensive dataset includes the complete history of 267 304 095 items covering entire Wikimedia projects from the onset to 2016. From this encyclopedic dataset, we observe universal interplay between number of edits and degree of inequality for a given communal datasets. Specifically, the rapid increasing of gini index suggests that this entrenchment inequality stem from the nature of such open-editing communal datasets, namely abiogenesis of "super-editors’ cartel." We present the evidence that these groups created at the early initial stage of these open-editing media and have kept alive until the present. Additionally our model regarding both short- and long- term memories successfully elucidates the mechanism to compose such unofficial government of Wikipedia. Eventually, our results forewarn the pessimistic prospect of such communal databases.