Signed networks in social media

By Jure Leskovec, Dan Huttenlocher, Jon M. Kleinberg. International Conference on Human Factors in Computing Systems, 2010

1,580 citations recorded by the source corpus.

Relations between users on social media sites often reflect a mixture of positive (friendly) and negative (antagonistic) interactions. In contrast to the bulk of research on social networks that has focused almost exclusively on positive interpretations of links between people, we study how the interplay between positive and negative relationships affects the structure of on-line social networks. We connect our analyses to theories of signed networks from social psychology. We find that the classical theory of structural balance tends to capture certain common patterns of interaction, but that it is also at odds with some of the fundamental phenomena we observe --- particularly related to the evolving, directed nature of these on-line networks. We then develop an alternate theory of status that better explains the observed edge signs and provides insights into the underlying social mechanisms. Our work provides one of the first large-scale evaluations of theories of signed networks using on-line datasets, as well as providing a perspective for reasoning about social media sites.

Subfields: Communication, Statistical and Nonlinear Physics

View Paper: https://www.semanticscholar.org/paper/aa3c0e570211d30b26121d5172cef27425ff71ad

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Communication

In Statistical and Nonlinear Physics

Citations in the corpus

Negative Link Prediction in Social Media (2014, background): “In this subsection, we investigate two of the most important social theories for signed networks, i.e., balance theory [9, 2] and status theory [14].”
Decentralized Federated Learning: Fundamentals, State of the Art, Frameworks, Trends, and Challenges (2022, methodology): “[73] 2022 Decentralized recommender system based on the principles of gossip learning Cross-device Decentralized Simulation GMF PRME-G FedAvg FedFast Reptile Synchronous Scheduling Foursqueare-NYC [178] Gowalla-NYC [179] MovieLens [180] [74] 2022 Gossip-based system to maximize social awareness amon”
Inferring social ties across heterogenous networks (2012, background): “We found near ly 99% of triads in the two networks satisfy the social status theory, which was also validated in [20].”
mTrust: discerning multi-faceted trust in a connected world (2012, background): “It is reported in [13] that the theory of status holds weakly on these networks that are reciprocated.”
Collaborative Filtering beyond the User-Item Matrix (2014, background): “However, the assumption that “an enemy’s enemy is a friend” is less reliable and of the negative relationships and its predictability [Guha et al. 2004; Leskovec et al. 2010b].”
FRAUDAR: Bounding Graph Fraud in the Face of Camouflage (2016, methodology): “We run FRAUDAR on four real-world graphs: Amazon [18], Trip Advisor [29], Epinions [17], and Wiki-vote [17].”
Structural balance in the social networks of a wild mammal (2013, background): “Weak, or generalized, structural balance is empirically more common than strong (regular) structural balance in human social networks (Leskovec et al. 2010; van de Rijt 2011), although recently, strong structural balance was demonstrated in a large human online network (Szell et al. 2010).”
Computational Sociolinguistics: A Survey (2016, background): “Work within this area has drawn from Structural Balance Theory [107], which captures intuitions such as that when two individuals have a mutual friend, they are likely to be friends as well, and from Status Theory [154],”
Computing global structural balance in large-scale signed social networks (2011, background): “As a matter of fact, for what concerns the online signed social networks currently available (see Materials and Methods for a description), only an analysis of local, low-dimensional motifs has been carried out so far (12, 13).”
DEBAGREEMENT: A comment-reply dataset for (dis)agreement detection in online debates (2021, methodology): “Modeling online discussion forums as graphs of interactions between users enables researchers to: i) translate the (dis)agreement detection task into a sign link prediction one [45], ii) use recent advances in GRL methods [20], and iii) leverage existing signed graph embedding methods [41, 15, 30] t”
Data Fusion by Matrix Factorization (2013, background): “2 or 3 would contain asymmetric or one-way relations, such as those from the analysis of signed networks [67] and computational biology [68], among others.”
Digital Footprints: Opportunities and Challenges for Online Social Research (2014, background): “Leskovec et al. (2010) examined tie formation in online networks including Epinions, Slashdot, and Wikipedia and found that undirected ties are formed as predicted by structural balance theory (the product of signs in a balanced triad must be positive), but when ties are directed, status effects…”
Hierarchical block structures and high-resolution model selection in large networks (2013, background): “com trust network [79] 28 Berkeley/Stanford web graph [71] 7 Power Grid [75] 18 Slashdot [80] 29 Google web graph [71] 8 arXiv Co-Authors (hep-th) [70] 19 Flickr [81] 30 Youtube social network [73] 9 arXiv Co-Authors (hep-ph) [70] 20 Gowalla social network [78] 31 Yahoo groupse (bipartite) 10 arXiv ”
The spreading of misinformation online (2016, background): “Because δ by itself cannot capture the homogeneous clusters observed in the data, we model the connectivity pattern as a signed network (4, 42) considering different fractions of homogeneous links and hence restricting diffusion of news only to homogeneous links.”
Clustering and Community Detection in Directed Networks: A Survey (2013, background): “[153, 154]) are trying to capture the notion of positive and negative interactions among the nodes of a network.”
Random walks and diffusion on networks (2016, background): “Status theory One can also understand a hierarchy in the context of “status theory”, in which certain nodes have a higher status than others, and a directed edge indicates a difference of status [175].”