Bayesian block modelling for weighted networks

Gallagher, Ian

doi:10.1145/1830252.1830260

Cited by 1 publication

(3 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Gallagher [10] further extended the stochastic blockmodel from binary graphs to weighted networks. In his model, the weight of an edge is the number of observations of such edge, e.g., the number of messages.…”

Section: Related Workmentioning

confidence: 99%

“…In the blockmodels proposed by Gallagher [10], the weight of a relationship is the count of interactions of some relation, e.g., the number of messages seen in an email network, which is modeled by a Poisson distribution. Here, we extend this blockmodel by considering multiple relations, i.e., there is one Poisson distribution associated with each relation for every role block.…”

Section: Gsbm With Poisson Distributions 41 Mpois: Basic Mvpdfmentioning

confidence: 99%

“…Here, we extend this blockmodel by considering multiple relations, i.e., there is one Poisson distribution associated with each relation for every role block. The blockmodel for a weighted single relation network in [10] now becomes a special case where each role block B uv is modeled by a single Poisson parameter λ uv .…”

Section: Gsbm With Poisson Distributions 41 Mpois: Basic Mvpdfmentioning

confidence: 99%

See 2 more Smart Citations

Structural Analysis in Multi-Relational Social Networks

Dai

Chua

Lim

2012

Proceedings of the 2012 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

Modern social networks often consist of multiple relations among individuals. Understanding the structure of such multi-relational network is essential. In sociology, one way of structural analysis is to identify different positions and roles using blockmodels. In this paper, we generalize stochastic blockmodels to Generalized Stochastic Blockmodels (GSBM) for performing positional and role analysis on multi-relational networks. Our GSBM generalizes many different kinds of Multivariate Probability Distribution Function (MVPDF) to model different kinds of multi-relational networks. In particular, we propose to use multivariate Poisson distribution for multi-relational social networks. Our experiments show that GSBM is able to identify the structures for both synthetic and real world network data. These structures can further be used for predicting relationships between individuals.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Gsbm With Poisson Distributions 41 Mpois: Basic Mvpdfmentioning

confidence: 99%

Section: Gsbm With Poisson Distributions 41 Mpois: Basic Mvpdfmentioning

confidence: 99%

See 1 more Smart Citation