Confident Network Indices with Latent Space Models
نویسندگان
چکیده
Although traditional social network analysis operates on the assumption that the observed relationships represents the true social network, this assumption is dangerous, especially in noisy environments. This assumption is especially problematic given the lack of robustness with respect to missing or erroneous information that has been found for node-level network indices such as degree centrality or betweenness centrality. We examine latent space models of network generation as one technique for estimating confidence in network indices. Using the Sampson monastery dataset, we show how latent space models can be used to construct confidence intervals for node-level network indices.
منابع مشابه
The Geometry of Continuous Latent Space Models for Network Data
We review the class of continuous latent space (statistical) models for network data, paying particular attention to the role of the geometry of the latent space. In these models, the presence/absence of network dyadic ties are assumed to be conditionally independent given the dyads? unobserved positions in a latent space. In this way, these models provide a probabilistic framework for embeddin...
متن کاملA latent space representation of overdispersed relative propensity in “How many X’s do you know?” data
We present a novel latent space representation of the relative propensity for a respondent to form ties with members of a particular social group, a quantity related to overdispersion. In many applications collecting complete network data is financially or practically infeasible. Instead, we use data where respondents are asked for the number of ties they have with members of various subpopulat...
متن کاملDecoupling Homophily and Reciprocity with Latent Space Network Models
Networks form useful representations of data arising in various physical and social domains. In this work, we consider dynamic networks such as communication networks in which links connecting pairs of nodes appear over continuous time. We adopt a point processbased approach, and study latent space models which embed the nodes into Euclidean space. We propose models to capture two different asp...
متن کاملLatent words recurrent neural network language models
This paper proposes a novel language modeling approach called latent word recurrent neural network language model, which solves the problems present in both recurrent neural network language models (RNNLMs) and latent word language models (LWLMs). The proposed model has a soft class structure based on a latent variable space as well as LWLM, where the latent variable space is modeled using RNNL...
متن کاملLatent space models for networks using Aggregated Relational Data1
Despite increasing interest across a range of scientific applications in modeling and understanding social network structure, collecting complete network data remains logistically and financially challenging, especially in the social sciences. This paper introduces a latent space representation of social network structure for partially observed network data. We derive a multivariate measure of ...
متن کامل