Recently, there has been tremendous interest in the phenomenon of influence propagation social networks. The studies this area assume they have as input to their problems a graph with edges labeled probabilities between users. However, question where these come from or how can be computed real network data largely ignored until now. Thus it is interesting ask whether and log actions by its users, one build models influence. This main problem attacked paper. In addition proposing algorithms...

Preserving individual privacy when publishing data is a problem that receiving increasing attention. According to the fc-anonymity principle, each release of must be such indistinguishable from at least k - 1 other individuals. In this paper we study anonymity preserving in moving objects databases. We propose novel concept k-anonymity based on co-localization exploits inherent uncertainty object's whereabouts. Due sampling and positioning systems (e.g., GPS) imprecision, trajectory object...

Influence maximization is the problem of finding a set users in social network, such that by targeting this set, one maximizes expected spread influence network. Most literature on topic has focused exclusively graph, overlooking historical data, i.e., traces past action propagations. In paper, we study from novel data-based perspective. particular, introduce new model, which call credit distribution , directly leverages available propagation to learn how flows network and uses estimate...

Social network analysis has gained significant attention in recent years, largely due to the success of online social networking and media-sharing sites, consequent availability a wealth data. In spite growing interest, however, there is little understanding potential business applications mining networks. While large body research on different problems methods for mining, gap between techniques developed by community their deployment real-world applications. Therefore impact these still...

Algorithms and decision making based on Big Data have become pervasive in all aspects of our daily lives (offline online), as they essential tools personal finance, health care, hiring, housing, education, policies. It is therefore societal ethical importance to ask whether these algorithms can be discriminative grounds such gender, ethnicity, or status. turns out that the answer positive: for instance, recent studies context online advertising show ads high-income jobs are presented men...

Query logs record the queries and actions of users search engines, as such they contain valuable information about interests, preferences, behavior users, well their implicit feedback to engine results. Mining wealth available in query has many important applications including query-log analysis, user profiling personalization, advertising, recommendation, more.In this paper we introduce query-flow graph, a graph representation interesting knowledge latent querying behavior. Intuitively,...

Finding dense subgraphs is an important graph-mining task with many applications. Given that the direct optimization of edge density not meaningful, as even a single achieves maximum density, research has focused on optimizing alternative functions. A very popular among such functions average degree, whose maximization leads to well-known densest-subgraph notion. Surprisingly enough, however, densest are typically large graphs, small and diameter.

In this paper we study approximate landmark-based methods for point-to-point distance estimation in very large networks. These involve selecting a subset of nodes as landmarks and computing offline the distances from each node graph to those landmarks. At runtime, when between pair is needed, it can be estimated quickly by combining precomputed distances. We prove that optimal set an NP-hard problem, thus heuristic solutions need employed. therefore explore theoretical insights devise...

In this work, we define and solve the Fair Top-k Ranking problem, in which want to determine a subset of k candidates from large pool n >> candidates, maximizing utility (i.e., select "best" candidates) subject group fairness criteria. Our ranked definition extends using standard notion protected groups is based on ensuring that proportion every prefix top-k ranking remains statistically above or indistinguishable given minimum. Utility operationalized two ways: (i) candidate included...

Complex networks, such as biological, social, and communication often entail uncertainty, thus, can be modeled probabilistic graphs . Similar to the problem of similarity search in standard graphs, a fundamental for is efficiently answer k-nearest neighbor queries ( k -NN), which computing closest nodes some specific node. In this paper we introduce framework processing -NN graphs. We propose novel distance functions that extend well-known graph concepts, shortest paths. order compute them...

We study social influence from a topic modeling perspective. introduce novel topic-aware influence-driven propagation models that experimentally result to be more accurate in describing real-world cascades than the standard studied literature. In particular, we first propose simple extensions of well-known Independent Cascade and Linear Threshold models. Next, different approach explicitly authoritativeness, relevance under devise methods learn parameters dataset past propagations. Our...

We introduce a novel frequent pattern mining approach to discover leaders and tribes in social networks. In particular, we consider networks where users perform actions. Actions may be as simple tagging resources (urls) del.icio.us, rating songs Yahoo! Music, or movies Movies, buying gadgets such cameras, handhelds, etc. blogging review on the gadgets. The assumption is that actions performed by user can seen their network friends. Users seeing friends' are sometimes tempted those interested...

User recommender systems are a key component in any on-line social networking platform: they help the users growing their network faster, thus driving engagement and loyalty.

In the traditional link prediction problem, a snapshot of social network is used as starting point to predict, by means graph-theoretic measures, links that are likely appear in future. this paper, we introduce cold start problem predicting structure when itself totally missing while some other information regarding nodes available. We propose two-phase method based on bootstrap probabilistic graph. The first phase generates an implicit under form second applies graph-based measures produce...

Core decomposition has proven to be a useful primitive for wide range of graph analyses. One its most appealing features is that, unlike other notions dense subgraphs, it can computed linearly in the size input graph. In this paper we provide an analogous tool uncertain graphs, i.e., graphs whose edges are assigned probability existence. The fact that core efficiently deterministic does not guarantee efficiency where even simplest operations may become computationally intensive. Here show...

Moving object databases (MOD) have gained much interest in recent years due to the advances mobile communications and positioning technologies. Study of MOD can reveal useful information (e.g., traffic patterns congestion trends) that be used applications for common benefit. In order mine and/or analyze data, must published, which pose a threat location privacy user. Indeed, based on prior knowledge user's at several time points, an attacker potentially associate user specific moving (MOB)...

We present Spine, an efficient algorithm for finding the "backbone" of influence network. Given a social graph and log past propagations, we build instance independent-cascade model that describes propagations. aim at reducing complexity model, while preserving most its accuracy in describing data.

Graph Evolution Rules help in analyzing the evolution of large networks and can be used to predict future creation links among nodes.

Every day millions of users are connected through online social networks, generating a rich trove data that allows us to study the mechanisms behind human interactions. Triadic closure has been treated as major mechanism for creating links: if Alice follows Bob and Charlie, will follow Charlie. Here we present an analysis longitudinal micro-blogging data, revealing more nuanced view strategies employed by when expanding their circles. While network structure affects spread information among...