Communicability across evolving networks

Many natural and technological applications generate time ordered sequences of networks, defined over a fixed set of nodes; for example time-stamped information about ‘who phoned who’ or ‘who came into contact with who’ arise naturally in studies of communication and the spread of disease. Concepts and algorithms for static networks do not immediately carry through to this dynamic setting. For example, suppose A and B interact in the morning, and then B and C interact in the afternoon. Information, or disease, may then pass from A to C, but not vice versa. This subtlety is lost if we simply summarize using the daily aggregate network given by the chain A-B-C. However, using a natural definition of a walk on an evolving network, we show that classic centrality measures from the static setting can be extended in a computationally convenient manner. In particular, communicability indices can be computed to summarize the ability of each node to broadcast and receive information. The computations involve basic operations in linear algebra, and the asymmetry caused by time’s arrow is captured naturally through the non-mutativity of matrix-matrix multiplication. Illustrative examples are given for both synthetic and real-world communication data sets. We also discuss the use of the new centrality measures for real-time monitoring and prediction.

A matrix iteration for dynamic network summaries

We propose a new algorithm for summarizing properties of large-scale time-evolving networks. This type of data, recording connections that come and go over time, is being generated in many modern applications, including telecommunications and on-line human social behavior. The algorithm computes a dynamic measure of how well pairs of nodes can communicate by taking account of routes through the network that respect the arrow of time. We take the conventional approach of downweighting for length (messages become corrupted as they are passed along) and add the novel feature of downweighting for age (messages go out of date). This allows us to generalize widely used Katz-style centrality measures that have proved popular in network science to the case of dynamic networks sampled at non-uniform points in time. We illustrate the new approach on synthetic and real data.

Interventions in social networks: impact on mood and network dynamics

Results from two studies on longitudinal friendship networks are presented, exploring the impact of a gratitude intervention on positive and negative affect dynamics in a social network. The gratitude intervention had been previously shown to increase positive affect and decrease negative affect in an individual but dynamic group effects have not been considered. In the first study the intervention was administered to the whole network. In the second study two social networks are considered and in each only a subset of individuals, initially low/high in negative affect respectively received the intervention as `agents of change'. Data was analyzed using stochastic actor based modelling techniques to identify resulting network changes, impact on positive and negative affect and potential contagion of mood within the group. The first study found a group level increase in positive and a decrease in negative affect. Homophily was detected with regard to positive and negative affect but no evidence of contagion was found. The network itself became more volatile along with a fall in rate of change of negative affect. Centrality measures indicated that the best broadcasters were the individuals with the least negative affect levels at the beginning of the study. In the second study, the positive and negative affect levels for the whole group depended on the initial levels of negative affect of the intervention recipients. There was evidence of positive affect contagion in the group where intervention recipients had low initial level of negative affect and contagion in negative affect for the group where recipients had initially high level of negative affect.

Centrality and spectral radius in dynamic communication networks

We explore the influence of the choice of attenuation factor on Katz centrality indices for evolving communication networks. For given snapshots of a network observed over a period of time, recently developed communicability indices aim to identify best broadcasters and listeners in the network. In this article, we looked into the sensitivity of communicability indices on the attenuation factor constraint, in relation to spectral radius (the largest eigenvalue) of the network at any point in time and its computation in the case of large networks. We proposed relaxed communicability measures where the spectral radius bound on attenuation factor is relaxed and the adjacency matrix is normalised in order to maintain the convergence of the measure. Using a vitality based measure of both standard and relaxed communicability indices we looked at the ways of establishing the most important individuals for broadcasting and receiving of messages related to community bridging roles. We illustrated our findings with two examples of real-life networks, MIT reality mining data set of daily communications between 106 individuals during one year and UK Twitter mentions network, direct messages on Twitter between 12.4k individuals during one week.

On the radius of centrality in evolving communication networks

In this article, we investigate how the choice of the attenuation factor in an extended version of Katz centrality influences the centrality of the nodes in evolving communication networks. For given snapshots of a network, observed over a period of time, recently developed communicability indices aim to identify the best broadcasters and listeners (receivers) in the network. Here we explore the attenuation factor constraint, in relation to the spectral radius (the largest eigenvalue) of the network at any point in time and its computation in the case of large networks. We compare three different communicability measures: standard, exponential, and relaxed (where the spectral radius bound on the attenuation factor is relaxed and the adjacency matrix is normalised, in order to maintain the convergence of the measure). Furthermore, using a vitality-based measure of both standard and relaxed communicability indices, we look at the ways of establishing the most important individuals for broadcasting and receiving of messages related to community bridging roles. We compare those measures with the scores produced by an iterative version of the PageRank algorithm and illustrate our findings with two examples of real-life evolving networks: the MIT reality mining data set, consisting of daily communications between 106 individuals over the period of one year, a UK Twitter mentions network, constructed from the direct \emph{tweets} between 12.4k individuals during one week, and a subset the Enron email data set.

Assessing reading at Key Stage 2: SATs as measures of children's inferential abilities

This article describes two studies. The first study was designed to investigate the ways in which the statutory assessments of reading for 11-year-old children in England assess inferential abilities. The second study was designed to investigate the levels of performance achieved in these tests in 2001 and 2002 by 11-year-old children attending state-funded local authority schools in one London borough. In the first study, content and questions used in the reading papers for the Standard Assessment Tasks (SATs) in the years 2001 and 2002 were analysed to see what types of inference were being assessed. This analysis suggested that the complexity involved in inference making and the variety of inference types that are made during the reading process are not adequately sampled in the SATs. Similar inadequacies are evident in the ways in which the programmes of study for literacy recommended by central government deal with inference. In the second study, scripts of completed SATs reading papers for 2001 and 2002 were analysed to investigate the levels of inferential ability evident in scripts of children achieving different SATs levels. The analysis in this article suggests that children who only just achieve the 'target' Level 4 do so with minimal use of inference skills. They are particularly weak in making inferences that require the application of background knowledge. Thus, many children who achieve the reading level (Level 4) expected of 11-year-olds are entering secondary education with insecure inference-making skills that have not been recognised.

Relation between the fatty acid composition of peripheral blood mononuclear cells and measures of immune cell function in health free-living subjects aged 25-72 y

Background: There is little information about the relation between the fatty acid composition of human immune cells and the function of those cells over the habitual range of fatty acid intakes. Objective: The objective of the study was to determine the relation between the fatty acid composition of human peripheral blood mononuclear cell (PBMC) phospholipids and the functions of human immune cells. Design: One hundred fifty healthy adult subjects provided a fasting blood sample. The phagocytic and oxidative burst activities of monocytes and neutrophils were measured in whole blood. PBMCs were isolated and used to measure lymphocyte proliferation in response to the T cell mitogen concanavalin A and the production of cytokines in response to concanavalin A or bacterial lipopolysaccharide. The fatty acid composition of plasma and PBMC phospholipids was determined. Results: Wide variations in fatty acid composition of PBMC phospholipids and immune cell functions were identified among the subjects. The proportions of total Polyunsaturated fatty acids (PUFAs), of total n-6 and n-3 PUFAs, and of several individual PUFAs in PBMC phospholipids were positively correlated with phagocytosis by neutrophils and monocytes, neutrophil oxidative burst, lymphocyte proliferation, and interferon gamma production. The ratios of saturated fatty acids to PUFAs and of n-6 to n-3 PUFAs were negatively correlated with these same immune functions. The relation of PBMC fatty acid composition to monocyte oxidative burst was the reverse of its relation to monocyte phagocytosis and neutrophil oxidative burst. Conclusion: Variations in the fatty acid composition of PBMC phospholipids account for some of the variability in immune cell functions among healthy adults.

Impact of menstrual cycle phase on insulin sensitivity measures and fasting lipids

There is evidence to suggest that insulin sensitivity may vary in response to changes in sex hormone levels. However, the results Of human studies designed to investigate changes in insulin sensitivity through the menstrual cycle have proved inconclusive. The aims of this Study were to 1) evaluate the impact of menstrual cycle phase on insulin sensitivity measures and 2) determine the variability Of insulin sensitivity measures within the same menstrual cycle phase. A controlled observational study of 13 healthy premenopausal women, not taking any hormone preparation and having regular menstrual cycles, was conducted. Insulin sensitivity (Si) and glucose effectiveness (Sg) were measured using an intravenous glucose tolerance test (IVGTT) with minimal model analysis. Additional Surrogate measures Of insulin sensitivity were calculated (homoeostasis model for insulin resistance [HOMA IR], quantitative insulin-to-glucose check index [QUICKI] and revised QUICKI [rQUICKI]), as well as plasma lipids. Each woman was tested in the luteal and follicular phases of her Menstrual cycle, and duplicate measures were taken in one phase of the cycle. No significant differences in insulin sensitivity (measured by the IVGTT or Surrogate markers) or plasma lipids were reported between the two phases of the menstrual cycle or between duplicate measures within the same phase. It was Concluded that variability in measures of insulin sensitivity were similar within and between menstrual phases.

Self-, parent-report and interview measures of obsessive-compulsive disorder in children and adolescents

Self-report measures of obsessive-compulsive disorder (OCD) in children and adolescents are needed for practical evaluation of severity and treatment response. We compared the self- and parent-report Obsessional Compulsive Inventory Revised (CHOCI-R) to the interview-based Child Yale-Brown Obsessive-Compulsive Scale (CY-BOCS) in a clinical sample of 285 children and adolescents with OCD. Classical test theory and item-response theory were applied to compare the instruments. The self- and parent-report CHOCI-R had good internal consistency and were strongly related to each other. The self- and parent-report CHOCI-R severity scores correlated with the CY-BOCS (Pearson's r 0.55 and 0.45 respectively). The CY-BOCS discriminated better at the severe end of the spectrum. The CHOCI-R provided better discrimination in the mild to moderate range. The time-efficient self- and parent-report alternatives will enable routine measurement of OCD severity in clinical practice. Estimates of equivalent summed scores are provided to facilitate comparison. (c) 2007 Elsevier Ltd. All rights reserved.