Efficient Star-based Truss Maintenance on Dynamic Graphs

Graphs are integral to model real-world complex systems like social networks, citation networks, transaction networks etc. Real-world graphs are mostly heterogeneous, continuously evolving dynamic graphs (i.e. Heterogeneous continuous-time dynamic graph - HCTDG). Modeling HCTDGs requires effective representation learning, which is difficult because of their entangled structural and temporal dependencies. We introduce HT-Graph, a novel framework aimed to improve link prediction in HCTDG graphs. HT-Graph addresses scalability and computational inefficiencies while focusing on enhancing link prediction accuracy. As formation of new links between nodes depends on their neighborhood, HT-Graph introduces a neighbor-aware memory module (i.e. memory module with a neighbor-store) that stores and updates local neighborhood information of each node efficiently for faster calculation of structural information, eliminating redundant computations required for traditional neighborhood sampling. To introduce parallelism, we used a neighbor-aware restarter to restart the training at any timestamp using interaction history. During training, the restarter module resets memory states at multiple timestamps and learns to mimic the encoder through the knowledge distillation process. This eliminates sequential dependencies, enabling HT-Graph to capture temporal dynamics and structural heterogeneity while ensuring scalability. HT-Graph outperforms state-of-the-art models in link prediction, providing higher scalability, efficiency, and better predictive performance even with limited data.

While Chaps. 5 and 6 addressed contextual factors of the music items themselves, this chapter targets the interaction between users and music items. We first describe how respective data can be organized in terms of abstraction level. Subsequently, we detail data acquisition from social networks, ratings, playlists, listening histories, and others. Furthermore, aspects of the user context are discussed, including the listener’s location and activity as well as time and weather, which are typically acquired through sensor data from smart phones and web services, respectively. Users’ connections in social networks are also discussed in this context. We round off the chapter by analyzing some factors of user intention when listening to music. How the discussed interaction traces and aspects of the user context and user intention can be exploited for music similarity and music recommendation is the topic of Chap. 8

Graphs are essential representations of many real-world data such as social networks. Recent years have witnessed the increasing efforts made to extend the neural network models to graph-structured data. These methods, which are usually known as the graph neural networks, have been applied to advance many graphs related tasks such as reasoning dynamics of the physical system, graph classification, and node classification. Most of the existing graph neural network models have been designed for static graphs, while many real-world graphs are inherently dynamic. For example, social networks are naturally evolving as new users joining and new relations being created. Current graph neural network models cannot utilize the dynamic information in dynamic graphs. However, the dynamic information has been proven to enhance the performance of many graph analytic tasks such as community detection and link prediction. Hence, it is necessary to design dedicated graph neural networks for dynamic graphs. In this paper, we propose DGNN, a new {\bf D}ynamic {\bf G}raph {\bf N}eural {\bf N}etwork model, which can model the dynamic information as the graph evolving. In particular, the proposed framework can keep updating node information by capturing the sequential information of edges (interactions), the time intervals between edges and information propagation coherently. Experimental results on various dynamic graphs demonstrate the effectiveness of the proposed framework.

BackgroundCommunity engagement is central to the conduct of health-related research studies as a way to determine priorities, inform study design and implementation, increase recruitment and retention, build relationships, and ensure that research meets the goals of the community. Community sensitization meetings, a form of community engagement, are often held prior to the initiation of research studies to provide information about upcoming study activities and resolve concerns in consultation with potential participants. This study estimated demographic, health, economic, and social network correlates of attendance at community sensitization meetings held in advance of a whole-population, combined behavioral, and biomedical research study in rural Uganda.Methods and findingsResearch assistants collected survey data from 1,630 adults participating in an ongoing sociocentric social network cohort study conducted in a rural region of southwestern Uganda. These community survey data, collected between 2016 and 2018, were linked to attendance logs from community sensitization meetings held in 2018 and 2019 before the subsequent community survey and community health fair. Of all participants, 264 (16%) attended a community sensitization meeting before the community survey, 464 (28%) attended a meeting before the community health fair, 558 (34%) attended a meeting before either study activity (survey or health fair), and 170 (10%) attended a meeting before both study activities (survey and health fair). Using multivariable Poisson regression models, we estimated correlates of attendance at community sensitization meetings. Attendance was more likely among study participants who were women (adjusted relative risk [ARR]health fair = 1.71, 95% confidence interval [CI], 1.32 to 2.21, p < 0.001), older age (ARRsurvey = 1.02 per year, 95% CI, 1.01 to 1.02, p < 0.001; ARRhealth fair = 1.02 per year, 95% CI, 1.01 to 1.02, p < 0.001), married (ARRsurvey = 1.74, 95% CI, 1.29 to 2.35, p < 0.001; ARRhealth fair = 1.41, 95% CI, 1.13 to 1.76, p = 0.002), and members of more community groups (ARRsurvey = 1.26 per group, 95% CI, 1.10 to 1.44, p = 0.001; ARRhealth fair = 1.26 per group, 95% CI, 1.12 to 1.43, p < 0.001). Attendance was less likely among study participants who lived farther from meeting locations (ARRsurvey = 0.54 per kilometer, 95% CI, 0.30 to 0.97, p = 0.041; ARRhealth fair = 0.57 per kilometer, 95% CI, 0.38 to 0.86, p = 0.007). Leveraging the cohort’s sociocentric design, social network analyses suggested that information conveyed during community sensitization meetings could reach a broader group of potential study participants through attendees’ social network and household connections. Study limitations include lack of detailed data on reasons for attendance/nonattendance at community sensitization meetings; achieving a representative sample of community members was not an explicit aim of the study; and generalizability may not extend beyond this study setting.ConclusionsIn this longitudinal, sociocentric social network study conducted in rural Uganda, we observed that older age, female sex, being married, membership in more community groups, and geographical proximity to meeting locations were correlated with attendance at community sensitization meetings held in advance of bio-behavioral research activities. Information conveyed during meetings could have reached a broader portion of the population through attendees’ social network and household connections. To ensure broader input and potentially increase participation in health-related research studies, the dissemination of research-related information through community sensitization meetings may need to target members of underrepresented groups.

Reproductive success is often highly skewed in animal populations. Yet the processes leading to this are not always clear. Similarly, connections in animal social networks are often nonrandomly distributed, with some individuals with many connections and others with few, yet whether there are simple explanations for this pattern has not been determined. Numerous social interactions involve dyads embedded within a wider network. As a result, it may be possible to model which individuals accumulate social interactions through a more general understanding of the social network's structure, and how this structure changes over time. We analysed fighting and mating interactions across the breeding season in a population of wild field crickets under surveillance from a network of video cameras. We fitted stochastic actor-oriented models to determine the dynamic process by which networks of cricket fighting and mating interactions form, and how they co-influence each other. We found crickets tended to fight those in close spatial proximity to them and those possessing a mutual connection in the fighting network, and heavier crickets fought more often. We also found that crickets that mated with many others tended to fight less in the following time period. This demonstrates that a mixture of spatial constraints, characteristics of individuals and characteristics of the immediate social environment are key for determining social interactions. The mating interaction network required very few parameters to understand its growth and thus its structure; only homophily by mating success was required to simulate the skew of mating interactions seen in this population. This demonstrates that relatively simple, but dynamic, processes can give highly skewed distributions of mating success. • Social network connections and patterns of mating success are both strongly skewed. • We modelled dynamic social (fighting) and mating networks in wild crickets. • The fighting network was influenced by proximity, sex, mass and mutual connections. • Only positive assortment by mating success was required to model the mating network. • In many populations, simple, dynamic processes may explain strong skews in fitness.

There are plenty of researches on group decision making (GDM) problem and most of them assume that all the experts are independent. However, the social network connection is an important characteristic among experts, and should been taken into account in the GDM decision process. The social network analysis (SNA) is a rapidly developing technology to deal with the problem about social network connections. In this paper, we address GDM problem with fuzzy preference relations where the experts have directed social network connections. And we use in-degree and out-degree centrality indices in SNA to build the social characteristics of the experts. Then we utilize the importance induced ordered weighted averaging(I-IOWA) operator to aggregate all the individual fuzzy preference relations with social indices. Furthermore,we investigate the reciprocity, additive consistency and acceptable consistency properties about the obtained group fuzzy preference relations. A procedure of addressing GDM problem with fuzzy preference relations where the experts are within directed social network connections is developed. Moreover, we compare the results of the case where all the experts are independent and the case where all the experts are within directed social network connections. Some useful tips that affect final outranking result in the directed graphical social network connection are obtained. In the end, we summarize the main results of this paper and point out some research directions for future research.

Prediction of links/ connections in social networks is useful for increasing business in the area of telecom and social media. All types of social networking organization are trying to increase their nodes, i.e. the number of members. So the calculation of link prediction is the most important task because this calculation will help how to increase the number of user on a social network. A new technique of link prediction, namely RSM index, is introduced in this paper. Few essential properties have been established. Also, a comparative study with the existing methods is depicted with suitable tables and graphs.

Homophily of music listening in online social networks of China

Understanding the evolution of dynamic network structures is an emerging and very interesting topic, which is motivated by several real applications in many scientific fields. In this article, we discuss the link prediction problem, which is one of the key issues in the analysis of social evolving networks. We propose a new hybrid approach to predict the connections in social networks. The approach is inspired from the particle swarm algorithm and is combined with supervised machine learning strategy into a hybrid system. The paper includes an experimental study using real world data sets to compare the proposed methods against other approaches. The obtained results show good performance and prove the effectiveness of the proposed method.

User community detection via embedding of social network structure and temporal content

DynHEN: A heterogeneous network model for dynamic bipartite graph representation learning

Graph clustering is a fundamental data mining task that clusters vertices into different groups. The structural graph clustering algorithm ( SCAN ) is a widely used graph clustering algorithm that derives not only clustering results, but also special roles of vertices like hubs and outliers. In this paper, we consider structural graph clustering on dynamic graphs under Jaccard similarity. The state-of-the-art index-based solution focuses on static graphs and incurs prohibitive update costs to maintain indices. Lately, an efficient approximate dynamic structural graph clustering algorithm Dyn-StrClu under Jaccard similarity is proposed. However, their solution needs to fix input parameters while parameter settings of SCAN usually need to be fine-tuned to achieve good clustering results. Motivated by these limitations, we present a study on devising effective index structures for SCAN algorithm on dynamic graphs. Similar to the state-of-the-art dynamic scheme, our main idea to reduce the time complexity is still by bringing approximation to clustering results. However, our solution does not need to fix the input parameters. To achieve this, our solution includes two key components. The first is to maintain a bottom- k sketch for each vertex so that the similarities of affected vertices can be easily updated. The second key is a bucketing strategy that allows us to update clustering results and roles of vertices efficiently. Our theoretical analysis shows that our proposed algorithm achieves O (log n · log M + m / pf ) expected update cost and guarantees to return approximate clustering results with probability 1 - pf after up to M updates. Extensive experiments show that our solution is up to two orders of magnitude faster than the state-of-the-art index-based solution while still achieving high-quality clustering results.

Graphs are the dominant formalism for modeling multi-agent systems. The algebraic connectivity of a graph is particularly important because it provides the convergence rates of consensus algorithms that underlie many multi-agent control and optimization techniques. However, sharing the value of algebraic connectivity can inadvertently reveal sensitive information about the topology of a graph, such as connections in social networks. Therefore, in this work we present a method to release a graph’s algebraic connectivity under a graph-theoretic form of differential privacy, called edge differential privacy. Edge differential privacy obfuscates differences among graphs’ edge sets and thus conceals the absence or presence of sensitive connections therein. We provide privacy with bounded Laplace noise, which improves accuracy relative to conventional unbounded noise. The private algebraic connectivity values are analytically shown to provide accurate estimates of consensus convergence rates, as well as accurate bounds on the diameter of a graph and the mean distance between its nodes. Simulation results confirm the utility of private algebraic connectivity in these contexts.

This study examined potentially bidirectional connections of older U.S. women's salivary testosterone with their social network connections. Data were from the 2005-2006 and 2010-2011 waves of the National Social Life, Health and Aging Project (NSHAP), a national probability sample of older U.S. adults. Autoregressive cross-lagged panel models tested linkages of women's testosterone with their social networks. Consistent with recent biological theory suggesting social modulation of hormones, a higher kin proportion in one's egocentric (person-centered) network, arguably a stable compositional feature, negatively predicted women's testosterone levels. In contrast, findings for tie strength were consistent with hormonal regulation of women's sociality-with both perceived support from friends and family, and closeness to network members, negatively influenced by testosterone. Rather than being a static and exogenous biological factor, older women's testosterone levels seem partly an outcome of their social context. Implications for sexual health and hormone therapy are discussed. However, this androgen also influences dimensions of their intimate networks critical to successful aging. Findings suggest the need for social scientists to engage with the neuroendocrine literature, which offers suggestions on linkages of hormones with specific network patterns.

ABSTRACTYoung Black gay/bisexual and other men who have sex with men (YB-GBMSM) are disproportionately impacted by HIV/AIDS. Novel intervention strategies are needed to optimize engagement in HIV care for this population. We sought to develop a group-level intervention to enhance resilience by augmenting social capital (defined as the sum of resources in an individual’s social network) among YB-GBMSM living with HIV, with the ultimate goal of improving engagement in HIV care. Our multiphase, community-based participatory research (CBPR) intervention development process included: (1) Development and maintenance of a youth advisory board (YAB) comprised of YB-GBMSM living with HIV; (2) Qualitative in-depth interviews with YB-GBMSM living with HIV; (3) Qualitative in-depth interviews with care and service providers at clinics and community-based organizations; and (4) Collaborative development of intervention modules and activities with our YAB, informed by social capital theory and our formative research results. The result of this process is Brothers Building Brothers By Breaking Barriers, a two-day, 10-module group-level intervention. The intervention does not focus exclusively on HIV, but rather takes a holistic approach to supporting youth and enhancing resilience. Intervention modules aim to develop resilience at the individual level (exploration of black gay identity, development of critical self-reflection and coping skills), social network level (exploring strategies for navigating family and intimate relationships) and community level (developing strategies for navigating clinical spaces and plans for community participation). Most intervention activities are interactive, in order to facilitate new social network connections – and accompanying social capital – within intervention groups. In summary, our intensive CBPR approach resulted in a novel, culturally-specific intervention designed to enhance HIV care engagement by augmenting resilience and social capital among YB-GBMSM living with HIV.