Friendship prediction and homophily in social media

Aiello, Luca Maria; Barrat, Alain; Schifanella, Rossano; Cattuto, Ciro; Markines, Benjamin; Menczer, Filippo

doi:10.1145/2180861.2180866

Cited by 315 publications

(175 citation statements)

References 55 publications

(48 reference statements)

Supporting

Mentioning

164

Contrasting

Unclassified

Order By: Relevance

“…Rad and Benyoucef [24] use the common group memberships of users in the YouTube social network to measure the similarities between them. Aiello et al [1] also observe a strong correlation between the social connectivity and the tagging/group participation behavior on social media platforms. Other researchers extract value homophily from the self-reported interest or user-defined tags on one's static profile ([14], [2]).…”

Section: B Value Homophily -Horizontal Measurementioning

confidence: 93%

Link Formation on Twitter: The Role of Achieved Status and Value Homophily

Sun¹,

Rui²

2017

Proceedings of the 50th Hawaii International Conference on System Sciences (2017)

View full text Add to dashboard Cite

Abstract-Homophily has been a widely recognized dominant factor in offline social network connection, which refers to one's propensity to seek interactions with others of similar status or values. Existing studies regarding homophily factors have been limited mostly to offline sociodemographic characteristics, such as race, gender, religion, education and occupation, which may not necessarily manifest homophily in online social network. Some researchers dabble in online social network, but they extract homophily characteristics from static user profile or link data, which has not incorporated the dynamic process of social network. To better understand the key factors in the establishment of online relationship, we explore a large data set on Twitter, which contains all initiated links by 1453 organizational Twitter users over three months. An initiated link refers to organization following a user who is currently not a follower of the organization. We crawl data on a daily basis and monitor whether the initiated one-way link ends up with a two-way relationship. Based on the established homophily theory, we define two online homophily factors: achieved status homophily (estimated by the gap of the followers count), value homophily (measured by the overlap ratio of common followee, Pearson correlation, and Cosine similarity between two users' tweets, respectively). We find that both homophily factors play a key role in the formation of online reciprocal relationship, and the effect of status homophily is larger for superior followee (one who has more followers than the corresponding organization) than for inferior followee (one who has less followers than the corresponding organization). Our finding not only extends the offline "individualindividual" homophily theory to the new online "organizationindividual" relationship, but also provides Twitter users insight into extending their social network by strategically targeting followee.

show abstract

Section: B Value Homophily -Horizontal Measurementioning

confidence: 93%

Link Formation on Twitter: The Role of Achieved Status and Value Homophily

Sun¹,

Rui²

2017

Proceedings of the 50th Hawaii International Conference on System Sciences (2017)

View full text Add to dashboard Cite

show abstract

“…Although the latter reason is inherent to the structure of the OSN and to the limit we impose on the number of crawled users, the former is essentially due to the privacy settings of the targets' friends and the OSN dynamics. Our results show that homophily in social networks [5,30] does not only allow us to infer hidden attributes of OSN users locally, but also allows us to efficiently navigate toward the target. Note that we do not assume any prior knowledge about the network structure and the users' distribution in the network.…”

Section: Introductionmentioning

confidence: 90%

“…Moreover, its privacy policies are notoriously designed to encourage public disclosure: the default policy for many important user attributes is everybody, i.e., full public visibility. 5 We also implemented our attack in Google+ in order to validate our findings in Facebook. This OSN is now the second largest OSN, after Facebook [40], and shares many privacy features with Facebook.…”

Section: Methodsmentioning

confidence: 99%

Nowhere to Hide: Navigating around Privacy in Online Social Networks

Humbert¹,

Studer²,

Grossglauser³

et al. 2013

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. In this paper, we introduce a navigation privacy attack, where an external adversary attempts to find a target user by exploiting publicly visible attributes of intermediate users. If such an attack is successful, it implies that a user cannot hide simply by excluding himself from a central directory or search function. The attack exploits the fact that most attributes (such as place of residence, age, or alma mater) tend to correlate with social proximity, which can be exploited as navigational cues while crawling the network. The problem is exacerbated by privacy policies where a user who keeps his profile private remains nevertheless visible in his friends' "friend lists"; such a user is still vulnerable to our navigation attack. Experiments with Facebook and Google+ show that the majority of users can be found efficiently using our attack, if a small set of attributes are known about the target as side information. Our results suggest that, in an online social network where many users reveal a (even limited) set of attributes, it is nearly impossible for a specific user to "hide in the crowd".

show abstract

“…In particular, it is widely applied in recommendation systems for information retrieval, helping search for new friends (Aiello et al 2012) and potential business collaborators (Akcora et al 2011;Mori et al 2012;Wu et al 2013), finding domain experts or co-authors in academic fields (Pavlov and Ichise 2007). Obviously, the concept of SNA models can be generalised.…”

Section: Introductionmentioning

confidence: 99%

Inter-variable correlation prediction with fuzzy connected-triples

Shang

Shen

2018

Soft Comput

View full text Add to dashboard Cite

Identification of hidden relationships between domain attributes from different data sources is of great practical significance and forms an emerging field in data mining. However, currently there seldom exist any systematic methods that can effectively handle this problem, especially when dealing with imprecisely described associations. In this paper, a novel data-driven approach for inter-variable correlation prediction is proposed by exploiting the concept of connected-triples. The work is implemented with the use of fuzzy logic. Through the exploitation of link strength measurements and fuzzy inference, the job of detecting similar or related variables can be accomplished via examining link relation patterns within and across different data sources. Empirical evaluation results are discussed, revealing the potential of the proposed work in predicting interesting attribute relations, while involving simple computation mechanisms.

show abstract

Friendship prediction and homophily in social media

Cited by 315 publications

References 55 publications

Link Formation on Twitter: The Role of Achieved Status and Value Homophily

Link Formation on Twitter: The Role of Achieved Status and Value Homophily

Nowhere to Hide: Navigating around Privacy in Online Social Networks

Inter-variable correlation prediction with fuzzy connected-triples

Contact Info

Product

Resources

About