Yana Volkovich

Core decomposition of uncertain graphs

Core decomposition has proven to be a useful primitive for a wide range of graph analyses. One of its most appealing features is that, unlike other notions of dense subgraphs, it can be computed linearly in the size of the input graph. In this paper we provide an analogous tool for uncertain graphs, i.e., graphs whose edges are assigned a probability of existence. The fact that core decomposition can be computed efficiently in deterministic graphs does not guarantee efficiency in uncertain graphs, where even the simplest graph operations may become computationally intensive. Here we show that core decomposition of uncertain graphs can be carried out efficiently as well. We extensively evaluate our definitions and methods on a number of real-world datasets and applications, such as influence maximization and task-driven team formation.

Communication Dynamics in Twitter During Political Campaigns: The Case of the 2011 Spanish National Election

The irruption of social media in the political sphere is generating repositories of “Big Data,” which can be mined to gain insights into communication dynamics. The research reported here relies on a large data set from Twitter to examine the activity, emotional content, and interactions of political parties and politicians during the campaign for the Spanish national elections in November 2011. The aim of this study is to investigate the adaptation of political parties to this new communication and organizational paradigm originating in the evolution of the Internet and online social networks. We analyze the reply and retweet networks of seven political parties with significant offline differences to assess their conversation and information diffusion patterns. We observe that political parties, and especially the major traditional parties, still tend to use Twitter just as a one-way flow communication tool. Moreover, we find evidence of a balkanization trend in the Spanish online political sphere, as observed in previous research for other countries.

On the Accuracy of Hyper-local Geotagging of Social Media Content

Social media users share billions of items per year, only a small fraction of which is geotagged. We present a data-driven approach for identifying non-geotagged content items that can be associated with a hyper-local geographic area by modeling the location distributions of n-grams that appear in the text. We explore the trade-off between accuracy and coverage of this method. Further, we explore differences across content received from multiple platforms and devices, and show, for example, that content shared via different sources and applications produces significantly different geographic distributions, and that it is preferred to model and predict location for items according to their source. Our findings show the potential and the bounds of a data-driven approach to assigning location data to short social media texts, and offer implications for all applications that use data-driven approaches to locate content.

In-Degree and PageRank of web pages: why do they follow similar power laws?

PageRank is a popularity measure designed by Google to rank Web pages. Experiments confirm that PageRank values obey a power law with the same exponent as In-Degree values. This paper presents a novel mathematical model that explains this phenomenon. The relation between PageRank and In-Degree is modeled through a stochastic equation, which is inspired by the original definition of PageRank, and is analogous to the well-known distributional identity for the busy period in the M/G/1 queue. Further, we employ the theory of regular variation and Tauberian theorems to prove analytically that the tail distributions of PageRank and In-Degree differ only by a multiplicative constant, for which we derive a closed-form expression. Our analytical results are in good agreement with experimental data.

Far from the eyes, close on the web: impact of geographic distance on online social interactions

Online friendship connections are often not representative of social relationships or shared interest between users, but merely provide a public display of personal identity. A better picture of online social behaviour can be achieved by taking into account the intensity of communication levels between users, yielding useful insights for service providers supporting this communication. Among the several factors impacting user interactions, geographic distance might be affecting how users communicate with their friends. While spatial proximity appears influencing how people connect to each other even on the Web, the relationship between social interaction and spatial distance remains unexplored. In this work we analyse the relationship between online user interactions and geographic proximity with a detailed study of a large Spanish online social service. Our results show that while geographic distance strongly affects how social links are created, spatial proximity plays a negligible role on user interactions. These findings offer new insights on the interplay between social and spatial factors influencing online user behaviour and open new directions for future research and applications.

Determining factors behind the PageRank log-log plot

We study the relation between PageRank and other parameters of information networks such as in-degree, out-degree, and the fraction of dangling nodes. We model this relation through a stochastic equation inspired by the original definition of PageRank. Further, we use the theory of regular variation to prove that PageRank and in-degree follow power laws with the same exponent. The difference between these two power laws is in a multiplicative constant, which depends mainly on the fraction of dangling nodes, average in-degree, the power law exponent, and the damping factor. The out-degree distribution has a minor effect, which we explicitly quantify. Finally, we propose a ranking scheme which does not depend on out-degrees.

Biographical social networks on Wikipedia: a cross-cultural study of links that made history

It is arguable whether history is made by great men and women or vice versa, but undoubtably social connections shape history. Analysing Wikipedia, a global collective memory place, we aim to understand how social links are recorded across cultures. Starting with the set of biographies in the English Wikipedia we focus on the networks of links between these biographical articles on the 15 largest language Wikipedias. We detect the most central characters in these networks and point out culture-related peculiarities. Furthermore, we reveal remarkable similarities between distinct groups of language Wikipedias and highlight the shared knowledge about connections between persons across cultures.

Jointly They Edit: Examining the Impact of Community Identification on Political Interaction in Wikipedia

Background In their 2005 study, Adamic and Glance coined the memorable phrase ‘divided they blog’, referring to a trend of cyberbalkanization in the political blogosphere, with liberal and conservative blogs tending to link to other blogs with a similar political slant, and not to one another. As political discussion and activity increasingly moves online, the power of framing political discourses is shifting from mass media to social media. Methodology/Principal Findings Continued examination of political interactions online is critical, and we extend this line of research by examining the activities of political users within the Wikipedia community. First, we examined how users in Wikipedia choose to display their political affiliation. Next, we analyzed the patterns of cross-party interaction and community participation among those users proclaiming a political affiliation. In contrast to previous analyses of other social media, we did not find strong trends indicating a preference to interact with members of the same political party within the Wikipedia community. Conclusions/Significance Our results indicate that users who proclaim their political affiliation within the community tend to proclaim their identity as a ‘Wikipedian’ even more loudly. It seems that the shared identity of ‘being Wikipedian’ may be strong enough to triumph over other potentially divisive facets of personal identity, such as political affiliation.

Extremal Dependencies and Rank Correlations in Power Law Networks

We analyze dependencies in complex networks characterized by power laws (Web sample, Wikipedia sample and a preferential attachment graph) using statistical techniques from the extreme value theory and the theory of multivariate regular variation. To the best of our knowledge, this is the first attempt to apply this well developed methodology to comprehensive graph data. The new insights this yields are striking: the three above-mentioned data sets are shown to have a totally different dependence structure between graph parameters, such as in-degree and PageRank. Based on the proposed approach, we suggest a new measure for rank correlations. Unlike most known methods, this measure is especially sensitive to rank permutations for top-ranked nodes. Using the new correlation measure, we demonstrate that the PageRank ranking is not sensitive to moderate changes in the damping factor.

Gender homophily in online dyadic and triadic relationships

Gender homophily, or the preference for interaction with individuals of the same gender, has been observed in many contexts, especially during childhood and adolescence. In this study we investigate such phenomenon by analyzing the interactions of the ∼10 million users of Tuenti, a Spanish social networking service popular among teenagers. In dyadic relationships we find evidence of higher gender homophily for women. We also observe a preference of users with more friends to connect to the opposite gender. A particularly marked gender difference emerges in signing up for the social networking service and adding the first friends, and in the interactions by means of wall messages. In these contexts we find evidence of a strong homophily for women, and little or no homophily for men. By examining the gender composition of triangle motifs, we observe a marked tendency of users to group into gender homogeneous clusters, with a particularly high number of male-only triangles. We show that age plays an important role in this context, with a tendency to higher homophily for young teenagers in both dyadic and triadic relationships. Our findings have implications for addressing gender gap issues, understanding adolescent online behavior and technology adoption, and modeling social networks.