scholarly journals Finding missing links in interaction networks

2019 ◽  
Author(s):  
J. Christopher D. Terry ◽  
Owen T. Lewis
Keyword(s):  
Network Structure ◽  
Research Effort ◽  
Structural Models ◽  
Ecological Communities ◽  
Bipartite Networks ◽  
Deficit Model ◽  
True Number ◽  
Underlying Network ◽  
Highly Correlated ◽  
Network Metrics

AbstractDocumenting which species interact within ecological communities is challenging and labour-intensive. As a result, many interactions remain unrecorded, potentially distorting our understanding of network structure and dynamics. We test the utility of four structural models and a new coverage-deficit model for predicting missing links in both simulated and empirical bipartite networks. We find they can perform well, but that the predictive power of structural models varies with the underlying network structure. Predictions can be improved by ensembling multiple models. Sample-coverage estimators of the number of missed interactions are highly correlated with the number of missed interactions, but strongly biased towards underestimating the true number of missing links. Augmenting observed networks with most-likely missing links improves estimates of qualitative network metrics. Tools to identify likely missing links can be simple to implement, allowing the prioritisation of research effort and more robust assessment of network properties.

scholarly journals Network Analysis for Economics and Finance: An Application to Firm Ownership

2021 ◽  
pp. 331-355
Author(s):  
Janina Engel ◽  
Michela Nardo ◽  
Michela Rancan
Keyword(s):  
Network Analysis ◽  
Network Structure ◽  
Giant Component ◽  
Model Data ◽  
Firm Ownership ◽  
Network Metrics ◽  
Different Levels ◽  
Significant Concentration

AbstractIn this chapter, we introduce network analysis as an approach to model data in economics and finance. First, we review the most recent empirical applications using network analysis in economics and finance. Second, we introduce the main network metrics that are useful to describe the overall network structure and characterize the position of a specific node in the network. Third, we model information on firm ownership as a network: firms are the nodes while ownership relationships are the linkages. Data are retrieved from Orbis including information of millions of firms and their shareholders at worldwide level. We describe the necessary steps to construct the highly complex international ownership network. We then analyze its structure and compute the main metrics. We find that it forms a giant component with a significant number of nodes connected to each other. Network statistics show that a limited number of shareholders control many firms, revealing a significant concentration of power. Finally, we show how these measures computed at different levels of granularity (i.e., sector of activity) can provide useful policy insights.

Continual representation learning for evolving biomedical bipartite networks

2021 ◽  
Author(s):  
Kishlay Jha ◽  
Guangxu Xun ◽  
Aidong Zhang
Keyword(s):  
Network Structure ◽  
Learning Strategy ◽  
Structure Learning ◽  
Fundamental Problem ◽  
Representation Learning ◽  
Research Area ◽  
Bipartite Network ◽  
Bipartite Networks ◽  
Straightforward Application ◽  
Low Dimensional

Abstract Motivation Many real-world biomedical interactions such as ‘gene-disease’, ‘disease-symptom’ and ‘drug-target’ are modeled as a bipartite network structure. Learning meaningful representations for such networks is a fundamental problem in the research area of Network Representation Learning (NRL). NRL approaches aim to translate the network structure into low-dimensional vector representations that are useful to a variety of biomedical applications. Despite significant advances, the existing approaches still have certain limitations. First, a majority of these approaches do not model the unique topological properties of bipartite networks. Consequently, their straightforward application to the bipartite graphs yields unsatisfactory results. Second, the existing approaches typically learn representations from static networks. This is limiting for the biomedical bipartite networks that evolve at a rapid pace, and thus necessitate the development of approaches that can update the representations in an online fashion. Results In this research, we propose a novel representation learning approach that accurately preserves the intricate bipartite structure, and efficiently updates the node representations. Specifically, we design a customized autoencoder that captures the proximity relationship between nodes participating in the bipartite bicliques (2 × 2 sub-graph), while preserving both the global and local structures. Moreover, the proposed structure-preserving technique is carefully interleaved with the central tenets of continual machine learning to design an incremental learning strategy that updates the node representations in an online manner. Taken together, the proposed approach produces meaningful representations with high fidelity and computational efficiency. Extensive experiments conducted on several biomedical bipartite networks validate the effectiveness and rationality of the proposed approach.

Restoration of pollination interactions in communities invaded by non-native plants.

2020 ◽  
pp. 377-390
Author(s):  
Christopher N. Kaiser-Bunbury ◽  
◽  
Benno I. Simmons ◽  
◽  
Keyword(s):  
Plant Species ◽  
Restoration Ecology ◽  
Invasive Plant ◽  
Ecosystem Processes ◽  
Community Level ◽  
Ecological Communities ◽  
Management Interventions ◽  
And Function ◽  
Network Metrics ◽  
The Impact

Invasive plant species degrade and homogenize ecosystems worldwide, thereby altering ecosystem processes and function. To mitigate and reverse the impact of invasive plants on pollination, a key ecosystem function, conservation scientists and practitioners restore ecological communities and study the impact of such management interventions on plant-pollinator communities. Here, we describe opportunities and challenges associated with restoring pollination interactions as part of a holistic ecosystem-based restoration approach. We introduce a few general concepts in restoration ecology, and outline best planning and evaluation practices of restoring pollination interactions on the community level. Planning involves the selection of suitable plant species to support diverse pollinator communities, which includes considerations of the benefits and disadvantages of using native vs exotic, and bridge and framework plant species for restoration. We emphasize the central role of scientific- and community-level approaches for the planning phase of pollination restoration. For evaluation purposes, we argue that appropriate network indicators have the advantage of detecting changes in species behaviour with consequences for ecosystem processes and functions before these changes show up in altered species communities. Suitable network metrics may include interaction diversity and evenness, and network measures that describe the distribution of species, such as network and species-level specialization, modularity and motifs. Finally, we discuss the usefulness of the network approach in evaluating the benefits of restoration interventions for pollination interactions, and propose that applied network ecologists take a central role in transferring theory into practice.

scholarly journals Inferring network properties based on the epidemic prevalence

2019 ◽  
Vol 4 (1) ◽  
Author(s):  
Long Ma ◽  
Qiang Liu ◽  
Piet Van Mieghem
Keyword(s):  
Time Series ◽  
Simulated Annealing ◽  
Level Of Detail ◽  
Dynamical Process ◽  
Underlying Network ◽  
Average Fraction ◽  
Network Properties ◽  
Open Question ◽  
Network Metrics ◽  
Network Type

Abstract Dynamical processes running on different networks behave differently, which makes the reconstruction of the underlying network from dynamical observations possible. However, to what level of detail the network properties can be determined from incomplete measurements of the dynamical process is still an open question. In this paper, we focus on the problem of inferring the properties of the underlying network from the dynamics of a susceptible-infected-susceptible epidemic and we assume that only a time series of the epidemic prevalence, i.e., the average fraction of infected nodes, is given. We find that some of the network metrics, namely those that are sensitive to the epidemic prevalence, can be roughly inferred if the network type is known. A simulated annealing link-rewiring algorithm, called SARA, is proposed to obtain an optimized network whose prevalence is close to the benchmark. The output of the algorithm is applied to classify the network types.

scholarly journals Learning Diverse Bayesian Networks

2019 ◽  
Vol 33 ◽  
pp. 7793-7800
Author(s):  
Cong Chen ◽  
Changhe Yuan
Keyword(s):  
Bayesian Networks ◽  
Bayesian Network ◽  
Network Structure ◽  
Search Algorithm ◽  
Noisy Data ◽  
Local Neighborhood ◽  
Causal Ordering ◽  
True Model ◽  
Underlying Network ◽  
Novel Method

Much effort has been directed at developing algorithms for learning optimal Bayesian network structures from data. When given limited or noisy data, however, the optimal Bayesian network often fails to capture the true underlying network structure. One can potentially address the problem by finding multiple most likely Bayesian networks (K-Best) in the hope that one of them recovers the true model. However, it is often the case that some of the best models come from the same peak(s) and are very similar to each other; so they tend to fail together. Moreover, many of these models are not even optimal respective to any causal ordering, thus unlikely to be useful. This paper proposes a novel method for finding a set of diverse top Bayesian networks, called modes, such that each network is guaranteed to be optimal in a local neighborhood. Such mode networks are expected to provide a much better coverage of the true model. Based on a globallocal theorem showing that a mode Bayesian network must be optimal in all local scopes, we introduce an A* search algorithm to efficiently find top M Bayesian networks which are highly probable and naturally diverse. Empirical evaluations show that our top mode models have much better diversity as well as accuracy in discovering true underlying models than those found by K-Best.

scholarly journals On the temporally flexible structure of plant-pollinator interaction networks

2020 ◽  
Author(s):  
Paul J. CaraDonna ◽  
Nickolas M. Waser
Keyword(s):  
Temporal Variation ◽  
Network Structure ◽  
Species Interactions ◽  
Interspecific Interactions ◽  
Activity Period ◽  
Individual Species ◽  
Ecological Communities ◽  
Short Term ◽  
Temporal Flexibility ◽  
Plant Pollinator

AbstractEcological communities consist of species that are joined in complex networks of interspecific interaction. The interactions that networks depict often form and dissolve rapidly, but this temporal variation is not well integrated into our understanding of the causes and consequences of network structure. If interspecific interactions exhibit temporal flexibility across time periods over which organisms co-occur, then the emergent structure of the corresponding network may also be temporally flexible, something that a temporally-static perspective would miss. Here, we use an empirical system to examine short-term flexibility in network structure (connectance, nestedness, and specialization), and in individual species interactions that contribute to that structure. We investigated weekly plant-pollinator networks in a subalpine ecosystem across three summer growing seasons. To link the interactions of individual species to properties of their networks, we examined weekly temporal variation in species’ contributions to network structure. As a test of the potential robustness of networks to perturbation, we also simulated the random loss of species from weekly networks. We then compared the properties of weekly networks to the properties of cumulative networks that aggregate field observations over each full season. A week-to-week view reveals considerable flexibility in the interactions of individual species and their contributions to network structure. For example, species that would be considered relatively generalized across their entire activity period may be much more specialized at certain times, and at no point as generalized as the cumulative network may suggest. Furthermore, a week-to-week view reveals corresponding temporal flexibility in network structure and potential robustness throughout each summer growing season. We conclude that short-term flexibility in species interactions leads to short-term variation in network properties, and that a season-long, cumulative perspective may miss important aspects of the way in which species interact, with implications for understanding their ecology, evolution, and conservation.

scholarly journals Modeling Network Evolution by Colored Petri Nets

2016 ◽  
Vol 15 (1) ◽  
pp. 41-61
Author(s):  
Suwimon VONGSINGTHONG ◽  
Sirapat BOONKRONG ◽  
Herwig UNGER
Keyword(s):  
Petri Nets ◽  
Network Structure ◽  
Information Diffusion ◽  
User Behavior ◽  
Network Evolution ◽  
Clustering Coefficient ◽  
Colored Petri Nets ◽  
Traffic Patterns ◽  
Underlying Network ◽  
Flow Of Information

Discovering how information was distributed was essential for tracking, optimizing, and controlling networks. In this paper, a premier approach to analyze the reciprocity of user behavior, content, network structure, and interaction rules to the interplay between information diffusion and network evolution was proposed. Parameterization and insight diffusion patterns were characterized based on the community structure of the underlying network using diffusion related behavior data, collected by a developed questionnaire. The user roles in creating the flow of information were stochastically modeled and simulated by Colored Petri Nets, where the growth and evolution of the network structure was substantiated through the formation of the clustering coefficient, the average path length, and the degree distribution. This analytical model could be used for various tasks, including predicting future user activities, monitoring traffic patterns of networks, and forecasting the distribution of content.

scholarly journals Public Opinion Early Warning Agent Model: A Deep Learning Cascade Virality Prediction Model Based on Multi-Feature Fusion

2021 ◽  
Vol 15 ◽  
Author(s):  
Liqun Gao ◽  
Yujia Liu ◽  
Hongwu Zhuang ◽  
Haiyang Wang ◽  
Bin Zhou ◽  
...  
Keyword(s):  
Deep Learning ◽  
Public Opinion ◽  
Prediction Model ◽  
Network Structure ◽  
Early Warning ◽  
Feature Fusion ◽  
Learning Model ◽  
Feature Model ◽  
Underlying Network ◽  
Deep Learning Model

With the rapid popularity of agent technology, a public opinion early warning agent has attracted wide attention. Furthermore, a deep learning model can make the agent more automatic and efficient. Therefore, for the agency of a public opinion early warning task, the deep learning model is very suitable for completing tasks such as popularity prediction or emergency outbreak. In this context, improving the ability to automatically analyze and predict the virality of information cascades is one of the tasks that deep learning model approaches address. However, most of the existing studies sought to address this task by analyzing cascade underlying network structure. Recent studies proposed cascade virality prediction for agnostic-networks (without network structure), but did not consider the fusion of more effective features. In this paper, we propose an innovative cascade virus prediction model named CasWarn. It can be quickly deployed in intelligent agents to effectively predict the virality of public opinion information for different industries. Inspired by the agnostic-network model, this model extracts the key features (independent of the underlying network structure) of an information cascade, including dissemination scale, emotional polarity ratio, and semantic evolution. We use two improved neural network frameworks to embed these features, and then apply the classification task to predict the cascade virality. We conduct comprehensive experiments on two large social network datasets. Furthermore, the experimental results prove that CasWarn can make timely and effective cascade virality predictions and verify that each feature model of CasWarn is beneficial to improve performance.

scholarly journals LifeWebs: A (global) database of bipartite ecological interaction networks

2021 ◽  
Vol 5 ◽  
Author(s):  
Philip Butterill ◽  
Leonardo Jorge ◽  
Shuang Xing ◽  
Tom Fayle
Keyword(s):  
Interaction Network ◽  
Community Level ◽  
Interaction Networks ◽  
Sampling Effort ◽  
Ecological Communities ◽  
Bipartite Networks ◽  
Data Types ◽  
User Friendliness ◽  
Global Database ◽  
Machine Readable

The structure and dynamics of ecological interactions are nowadays recognized as a crucial challenge to comprehend the assembly, functioning and maintenance of ecological communities, their processes and the services they provide. Nevertheless, while standards and databases for information on species occurrences, traits and phylogenies have been established, interaction networks have lagged behind on the development of these standards. Here, we discuss the challenges and our experiences in developing a global database of bipartite interaction networks. LifeWebs*1 is an effort to compile community-level interaction networks from both published and unpublished sources. We focus on bipartite networks that comprise one specific type of interaction between two groups of species (e.g., plants and herbivores, hosts and parasites, mammals and their microbiota), which are usually presented in a co-occurrence matrix format. However, with LifeWebs, we attempt to go beyond simple matrices by integrating relevant metadata from the studies, especially sampling effort, explicit species information (traits and taxonomy/phylogeny), and environmental/geographic information on the communities. Specifically, we explore 1) the unique aspects of community-level interaction networks when compared to data on single inter-specific interactions, occurrence data, and other biodiversity data and how to integrate these different data types. 2) The trade-off between user friendliness in data input/output vs. machine-readable formats, especially important when data contributors need to provide large amounts of data usually compiled in a non-machine-readable format. 3) How to have a single framework that is general enough to include disparate interaction types while retaining all the meaningful information. We envision LifeWebs to be in a good position to test a general standard for interaction network data, with a large variety of already compiled networks that encompass different types of interactions. We provide a framework for integration with other types of data, and formalization of the data necessary to represent networks into established biodiversity standards.

Sign in / Sign up

Export Citation Format

Share Document