Analysis of Link Formation, Persistence and Dissolution in NetSense Data

📝 Abstract

We study a unique behavioral network data set (based on periodic surveys and on electronic logs of dyadic contact via smartphones) collected at the University of Notre Dame.The participants are a sample of members of the entering class of freshmen in the fall of 2011 whose opinions on a wide variety of political and social issues and activities on campus were regularly recorded - at the beginning and end of each semester - for the first three years of their residence on campus. We create a communication activity network implied by call and text data, and a friendship network based on surveys. Both networks are limited to students participating in the NetSense surveys. We aim at finding student traits and activities on which agreements correlate well with formation and persistence of links while disagreements are highly correlated with non-existence or dissolution of links in the two social networks that we created. Using statistical analysis and machine learning, we observe several traits and activities displaying such correlations, thus being of potential use to predict social network evolution.

💡 Analysis

We study a unique behavioral network data set (based on periodic surveys and on electronic logs of dyadic contact via smartphones) collected at the University of Notre Dame.The participants are a sample of members of the entering class of freshmen in the fall of 2011 whose opinions on a wide variety of political and social issues and activities on campus were regularly recorded - at the beginning and end of each semester - for the first three years of their residence on campus. We create a communication activity network implied by call and text data, and a friendship network based on surveys. Both networks are limited to students participating in the NetSense surveys. We aim at finding student traits and activities on which agreements correlate well with formation and persistence of links while disagreements are highly correlated with non-existence or dissolution of links in the two social networks that we created. Using statistical analysis and machine learning, we observe several traits and activities displaying such correlations, thus being of potential use to predict social network evolution.

📄 Content

2016 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM)

1197

Analysis of Link Formation, Persistence and Dissolution in NetSense Data

Ashwin Bahulkar1 and Boleslaw K. Szymanski1,2 1Department of Computer Science & NEST Center Rensselaer Polytechnic Institute
Troy, NY, USA 2Społeczna Akademia Nauk Łódź, Poland {bahula,szymab}@rpi.edu Omar Lizardo3,5, Yuxiao Dong4, Yang Yang4, and Nitesh V. Chawla4,5 3Department of Sociology & iCeNSA 4Computer Science & Engineering Departments & iCeNSA University of Notre Dame, Notre Dame, IN, USA 5Wrocław University of Technology, Wrocław, Poland {olizardo,ydong1,yyang1,nchawla}@nd.edu

Abstract— We study a unique behavioral network data set (based on periodic surveys and on electronic logs of dyadic contact via smartphones) collected at the University of Notre Dame. The participants are a sample of members of the entering class of freshmen in the fall of 2011 whose opinions on a wide variety of political and social issues and activities on campus were regularly recorded—at the beginning and end of each semester— for the first three years of their residence on campus. We create a communication activity network implied by call and text data, and a friendship network based on surveys. Both networks are limited to students participating in the NetSense surveys. We aim at finding student traits and activities on which agreements correlate well with formation and persistence of links while disagreements are highly correlated with non-existence or dissolution of links in the two social networks that we created. Using statistical analysis and machine learning, we observe several traits and activities displaying such correlations, thus being of potential use to predict social network evolution. Keywords—NetSense; social networks; evolving networks; link prediction; link persistence I. INTRODUCTION Renewed attention to dynamics in social network analysis and network science in the recent literature has brought back a concern with classical questions regarding the origins of personal relationships [2, 7], as well as the factors that account for their temporal persistence [8, 9, 11]. This has re-opened central issues under-emphasized in classical network theory: the problem of the emergence of network ties, the problem of the evolution of social relationships over time and the factors that contribute to dynamic tie persistence and decay. Standard contagion-based models propose that persons become more similar because they share a social tie [14]. From this perspective networks evolve to behavioral commonality via influence-based processes. The key assumption is that the network itself evolves independently of behavior and attitudes. In contrast to this assumption, network co-evolution models suggest that in the very same way in which network ties may result in the strengthening or weakening of behavioral propensities, it is also possible that previously existing agreement on certain behavioral propensities will be responsible for the strengthening or weakening of network ties. Social relationships have effects on behavior and attitudes, but behavior and attitudes may also have an effect on the structure of social relationships.

A key problem in empirically investigating this side of the co-evolution process is that agreement (or disagreement) on behavioral propensities may affect social ties via two distinct mechanisms. First, previous agreement may generate new ties were none previously existed. This process has been referred as (status, value, or choice) “homophily” in classical [2] and contemporary social network theory [11, 12]. Here, pre- existing behavioral and attitudinal commonalities facilitate the tie formation process. Second, as has been noted by other analysts agreements may also affect network dynamics via a pruning or negative selection process, whereby existing disagreements either prevent new ties from forming or lead to faster decay of low agreement ties in relation to high agreement ties [9].
With data taken at a single point in time, it is impossible to tease apart whether network evolution is driven by homophily or selective decay. This leads to threats to causal inference or to over-inflated estimates of influence and contagion processes [9, 13]. In this study, we leverage unique data containing over- time information on opinion, attitudinal, and behavioral agreement, as well as unobtrusively collected data on electronic communication to establish whether network dynamics are more deeply affected by value homophily, unfriending dynamics or both at the same time. II. NETSENSE DATA The NetSense data that used in this study consists of students reporting their on-campus activities, personal traits and interests, as well as views and opinions on various social issues at the beginning of every school semester from t

This content is AI-processed based on ArXiv data.