A new study, which has analyzed around 9 billion phone calls throughout almost a one-year period, is the first to identify details of features of the communication process and to quantify their impact in the diffusion of information.
"This is something very important in the processes such as the diffusion of commercial information, viral marketing and the market trends of products, but also in situations such as the spreading of rumours, opinions, policies, etc.," explained one of the authors of this research, Esteban Moro, from the Grupo de Investigación Interdisciplinar de Sistemas Complejos (Interdisciplinary Research Group of Complex Systems) of the Mathematics Department at the UC3M and the Instituto de Ciencias Matemáticas ICMAT (Institute of Mathematic Sciences), a CSIC, UAM, UCM and UC3M joint center.
The study's main conclusion, published in the journal Physical Review E, is the finding that people communicate in bursts. In this way, our behaviour (in communication as well as in other activities) does not happen in a homogenous way over time, but rather there is universal behaviour in which there is no communication, followed by short intervals called bursts. "This aspect of human activity which has also been observed in other activities such as e-mail, web page visits and stock market operations governs communication between people," , the researchers concluded. The effect of the bursts is that it slows down the information diffusion since the large periods of inactivity in the communication between two persons make it less likely that information is passed from one to the other. The study also highlights another important aspect of human communications: in group conversations, that is, although it is produced in bursts, these bursts happen at the same time among the members of the social group, which then accelerates information diffusion within these groups.
When, how much, and how do people communicate
The main objective of the research is to try to understand the temporal pattern of communication among persons in a social network. "As opposed to the static vision of a social network (who I interact with), our study seeks to understand when and how these social relations are produced," Professor Esteban Moro elaborated. And with two purposes, the first being, to see if relations in a social network can be quantified better, that is, determine if the communication rhythm between two persons allows us to know something about the "strength" or characteristics of the relation (family member, acquaintance, friend, colleague, etc..); and the second, to investigate the impact of these rhythms on the propagation of information in social networks, in processes such as viral marketing, product recommendations, etc.
This study, which receives funding from the Ministry of Science and Innovation through the MOSAICO Science and Innovation project, has been possible thanks to the framework agreement that the UC3M has with Telefónica I+D and a program of Becas de Formación de Doctores. (Professional Training Grants for PhD's) In addition, the company facilitated access to totally anonymous databases of calls in a country during a period of eleven months. In all, the researchers analyzed approximately 9,000 million calls between 20 million people, around 30 percent of the population of a country. "This large volume of data and the extended time period guarantees the representiveness and universality in the results," pointed out another one of the study's author, Giovanna Miritello, a PhD candidate in the Mathematics Department of UC3M, with a grant-contract from Telefónica I+D.
Materials provided by Universidad Carlos III de Madrid - Oficina de Información Científica. Note: Content may be edited for style and length.
Cite This Page: