Multichannel Social Signatures and Persistent Features of Egocentric Networks
Loading...
URL
Journal Title
Journal ISSN
Volume Title
Perustieteiden korkeakoulu |
Master's thesis
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.
Authors
Date
2018-05-14
Department
Major/Subject
Complex Systems
Mcode
SCI3060
Degree programme
Master’s Programme in Life Science Technologies
Language
en
Pages
45+6
Series
Abstract
Mobile phones are perfect sensors for capturing the behavior of people. They are widespread personal devices that we carry around all day. Modern smartphones, equipped with an arsenal of various sensors, monitor their environments and also their owners. However, even the simplest mobile phone device, when used with a SIM card, can collect rich behavioral data. Call Detail Records (CDRs), collected by telecommunication companies for billing purposes, contain detailed information on communication behavior of the users which can not be collected by traditional data collection methods such as questionnaires. Scientists have used CDRs to study the structure and dynamics of societal-level communication networks as well as the properties of egocentric networks. The structure of weighted egocentric networks can be quantified with the so-called social signatures. It is known that call-based social signatures are distinct and persistent at the individual level. However, calling is just one of the several channels that people use to communicate. To get a more realistic picture of people's social behavior we should include more communication channels. However, because of their intrinsic differences, it is challenging to combine the usage frequencies on multiple channels into single combined weights. In this Thesis, we propose a method for determining link weights which enables us to compare the egocentric networks across different channels and also to construct multichannel egocentric networks and multichannel social signatures. Using two different datasets on calling and texting behavior of people, we observed that similarly to call signatures, text-message signatures and multichannel signatures (combining information on calls and texts) are also persistent in time. Moreover, we observed that even though people call and text different sets of people, their call and text signatures are similar in shape. In other words, the shapes of our social signatures--which are distinct from signatures of others--seem to be independent of the communication channel or the people whom we contact. Further research is needed to explain the mechanism behind these shapes and to investigate the roots of persistence and stability of social signatures.Description
Supervisor
Saramäki, JariThesis advisor
Saramäki, JariKeywords
social networks, social signatures, egocentric networks, inferring social ties