Search

Interaction networks are a widely used tool to understand the dynamics of plant–pollinator ecological communities. However, while most mutualistic networks have been defined at the species level, ecological processes such as pollination take place at different scales, including the individual or patch levels. Yet, current approaches studying fine‐grain sub‐specific plant–pollinator networks only account for interactions among nodes belonging to a single plant species due to the conceptual and mathematical limitations of modeling simultaneously several plant species each composed of several nodes. Here, we introduce a multilayer diffusion network framework that allows modeling simple diffusion processes between nodes pertaining to the same or different layers (i.e. species). It is designed to depict from the network structure the potential conspecific and heterospecific pollen flows among plant individuals or patches. This potential pollen flow is modeled as a transport‐like system, in which pollen grain movements are represented as random‐walkers that diffuse on an ensemble of bipartite layers of conspecific plants and their shared pollinators. We exemplify this physical conceptualization using a dataset of nine fine‐grain sub‐specific plant–pollinator networks from a Mediterranean grassland of annual plants, where plant nodes represent groups of conspecifics within patches of 1 m2. The diffusion networks show pollinators effectively connecting sets of patches of the same and different plant species, forming a modular structure. Interestingly, different properties of the network structure, such as the conspecific pollen arrival probability and the number of conspecific subgraphs in which plants are embedded, are critical for the seed production of different plant species. We provide a simple but robust set of metrics to calculate potential pollen flow and scale down network ecology to functioning properties at the individual or patch level, where most ecological processes take place, hence moving forward the description and interpretation of species‐rich communities across scales. [ABSTRACT FROM AUTHOR]

Abstract To measure node importance, network scientists employ centrality scores that typically take a microscopic or macroscopic perspective, relying on node features or global network structure. However, traditional centrality measures such as degree centrality, betweenness centrality, or PageRank neglect the community structure found in real-world networks. To study node importance based on network flows from a mesoscopic perspective, we analytically derive a community-aware information-theoretic centrality score based on network flow and the coding principles behind the map equation: map equation centrality. Map equation centrality measures how much further we can compress the network’s modular description by not coding for random walker transitions to the respective node, using an adapted coding scheme and determining node importance from a network flow-based point of view. The information-theoretic centrality measure can be determined from a node’s local network context alone because changes to the coding scheme only affect other nodes in the same module. Map equation centrality is agnostic to the chosen network flow model and allows researchers to select the model that best reflects the dynamics of the process under study. Applied to synthetic networks, we highlight how our approach enables a more fine-grained differentiation between nodes than node-local or network-global measures. Predicting influential nodes for two different dynamical processes on real-world networks with traditional and other community-aware centrality measures, we find that activating nodes based on map equation centrality scores tends to create the largest cascades in a linear threshold model.

Background: Total-RNA sequencing (total-RNA-seq) allows the simultaneous study of both the coding and the non-coding transcriptome. Yet, computational pipelines have traditionally focused on particular biotypes, making assumptions that are not fullfilled by total-RNA-seq datasets. Transcripts from distinct RNA biotypes vary in length, biogenesis, and function, can overlap in a genomic region, and may be present in the genome with a high copy number. Consequently, reads from total-RNA-seq libraries may cause ambiguous genomic alignments, demanding for flexible quantification approaches. Results: Here we present Multi-Graph count (MGcount), a total-RNA-seq quantification tool combining two strategies for handling ambiguous alignments. First, MGcount assigns reads hierarchically to small-RNA and long-RNA features to account for length disparity when transcripts overlap in the same genomic position. Next, MGcount aggregates RNA products with similar sequences where reads systematically multi-map using a graph-based approach. MGcount outputs a transcriptomic count matrix compatible with RNA-sequencing downstream analysis pipelines, with both bulk and single-cell resolution, and the graphs that model repeated transcript structures for different biotypes. The software can be used as a python module or as a single-file executable program. Conclusions: MGcount is a flexible total-RNA-seq quantification tool that successfully integrates reads that align to multiple genomic locations or that overlap with multiple gene features. Its approach is suitable for the simultaneous estimation of protein-coding, long non-coding and small non-coding transcript concentration, in both precursor and processed forms. Both source code and compiled software are available at https://github.com/hitaandrea/MGcount. [ABSTRACT FROM AUTHOR]

A systemically important firm could be too-connected-to-fail and/or too-important-to-fail,two properties which centrality measures and community detection methods can capture respectively.This paper examines the performance of these measures in a variance decomposition global financialnetwork. Too-connected-to-fail risk and vulnerability rankings are quite robust to the choice ofcentrality measure. The PageRank centrality measure, however, does not seem as suitable for assessingvulnerabilities. Two community identification methods, edge betweenness and the map equation(Infomap) were used to identify systemic communities, which in turn capture the too-important-tofaildimension of systemic risk. The first method appears more robust to di erent weighting schemesbut tends to isolate too many firms. The second method exhibits the opposite characteristics. Overall,the analysis suggests that centrality measures and community identification methods complement eachother for assessing systemic risk in financial networks.

Aim: In recent years, novel approaches have been proposed to improve current bioregionalization methods, but these have not been thoroughly compared. We assessed the applicability of the recently developed network‐based clustering method (Infomap algorithm) in bioregionalization analysis at regional spatial scales and compared the results with commonly used distance‐based methods (hierarchical clustering algorithm). We also identified climate regions by using a model‐based cluster analysis (Gaussian algorithm). Finally, we quantified the representation of climate regions and bioregions in current protected areas (PAs). Location: Iran. Taxa: Terrestrial mammals. Methods: To define bioregions we used the Infomap algorithm and distance‐based clustering methods based on species distribution data (over 14,000 occurrence records for 188 species). The Infomap algorithm was applied using the interactive web application "INFOMAP BIOREGIONS" and the distance‐based clustering was based on unweighted pair‐group method using arithmetic averages (UPGMA). To identify climate regions we used principal components analysis and a model‐based cluster analysis both based on 15 climatic variables as well as a terrain ruggedness index. Results: The Infomap algorithm detected nine biogeographical units: seven bioregions and two transition zones. The distance‐based method suggested five bioregions. The identified bioregions differed between methods with some consistent spatial patterns across methods. Temperature and precipitation explained 85.8% of the environmental variation. Eight climate regions were identified. In general, climate variation and bioregional patterns are currently poorly represented in PAs (<25% coverage). Main conclusions: The network‐based method allowed identifying bioregions at regional scale and was apparently more sensitive than the conventional distance‐based method. The detection of transition zones by the Infomap algorithm was an advantage, and stressed the fact that the distribution of Iranian mammalian fauna is complex, especially in the south‐eastern part where contact areas between several bioregions are found. The identified bioregions (especially the distance‐based bioregions) and climate regions tended to match well with previous bioregionalization studies and the global terrestrial ecoregions. When thoroughly compared and understood, bioregions and climate regions provide a framework for regional biodiversity conservation planning. [ABSTRACT FROM AUTHOR]

The interactive effect is significant in the Chinese stock market, exacerbating the abnormal market volatilities and risk contagion. Based on daily stock returns in the Shanghai Stock Exchange (SSE) A-shares, this paper divides the period between 2005 and 2018 into eight bull and bear market stages to investigate interactive patterns in the Chinese financial market. We employ the Least Absolute Shrinkage and Selection Operator (LASSO) method to construct the stock network, compare the heterogeneity of bull and bear markets, and further use the Map Equation method to analyse the evolution of modules in the SSE A-shares market. Empirical results show that (1) the connected effect is more significant in bear markets than bull markets and gives rise to abnormal volatilities in the stock market; (2) a system module can be found in the network during the first four stages, and the industry aggregation effect leads to module differentiation in the last four stages; (3) some stocks have leading effects on others throughout eight periods, and medium- and small-cap stocks with poor financial conditions are more likely to become risk sources, especially in bear markets. Our conclusions are beneficial to improving investment strategies and making regulatory policies.

• Face clustering is first reformulated as non-overlapping community detection. • Three new metrics for facial image clustering are introduced. • A novel framework achieves new state-of-the-arts in face clustering. [Display omitted] Face clustering is a critical task in computer vision due to the increasing number of applications such as augmented reality or photo album management. The primary challenge in this task arises from the imperfections in image feature representations. Given image features extracted from an existing pre-trained representation model, it remains an unresolved problem that how to leverage the inherent characteristics of similarities among unlabelled images to improve the clustering performance. In order to solve face clustering in an unsupervised manner, we develop an effective and robust framework named as Adapt-Infomap. First, we reformulate face clustering as a process of non-overlapping community detection. Specially, Adapt-Infomap achieves face clustering by minimizing the entropy of information flows (also known as the map equation) on an affinity graph of images. Since the affinity graph of images might contain noisy edges, we develop an outlier detection strategy in Adapt-Infomap to adaptively refine the affinity graph. Experiments with ablation studies demonstrate that Adapt-Infomap significantly outperforms existing methods and achieves new state-of-the-arts on three popular large-scale datasets for face clustering, e.g. , an absolute improvement of more than 10 % and 3 % comparing with prior unsupervised and supervised methods respectively in terms of average of Pairwise F-score. [ABSTRACT FROM AUTHOR]

AI-related ethical issues, such as privacy leaks (Srivastava, 2013), algorithm discrimination (Žliobaitė, 2017), have an enormous impact on public life. Current research mostly focuses on the establishment and elaboration of guidelines of AI ethics (Jobin et al, 2019), which, however, are often too vague to implement. An in-depth analysis of public opinions about AI ethics is required in order to formulate actionable guidelines. This study employs the community detection to identify the hierarchical structure of knowledge about AI ethics, and explores the relationship between knowledge communities, using Wikipedia as a collaborative data source with sufficient cross-validation (Stephany, 2017). It aims to analyze the status quo of the public cognition of AI ethics with the community-based network, and provide the necessary support and reference for AI ethics policy makers. The workflow of this study is shown in Figure 1. First, we used the entry “Ethics of Artificial Intelligence” as the source link to crawl the Wikipedia data. The collected dataset is expressed as unweighted directed graph. Then, the Infomap algorithm (Rosvall and Bergstrom 2008, Leskovec et al. 2010) was applied to detect hidden structures in the network, which combined community detection and information coding. The core idea of Infomap algorithm is to use the shortest codelength to describe the path generated by a random walk. Two types of information codes were required for coding – codes of different groups and codes of different nodes within a group. A shorter encoding length means a better community partition, and better results of community detection. Large networks contain plentiful information, so how to extract information from large network structures is an important task of visualization. We chose the method of map equation to find, evaluate, and visualize the modular organization of the network (Rosvall, 2009). The map equation framework is flexible and can be combined with the Infomap search algorithm to identify two-level, multilevel, and overlapping organizations in weighted, directed, and multiplex networks (Bohlin et al, 2014). In the end, we analyzed the visualization to explore the hierarchical community structure of AI ethics. Some preliminary results are shown in Fig. 2. Fig. 2(a) is the network of the original data of Wikipedia, which is a direct display of crawler data. The red dot is seed link, namely ethics of artificial intelligence. We can get a gradually extend network from this dot, and the figure displays the labels of several important nodes. Fig. 2(b) shows the communities detected by Infomap. Different colors represent different communities which can be explicitly labelled if desired. Interactive visual analytical methods can then be applied to extract topics from the individual communities and explore their relationships. This may eventually lead to the creation of a hierarchical knowledge structure of AI ethics, allowing us to understand the main risks that raise public concerns about AI ethics, and enforce actionable AI ethics guidelines. [ABSTRACT FROM AUTHOR]

There are plans to start building offshore marine renewable energy devices throughout the Mediterranean and the Adriatic has been identified as a key location for wind farm developments. The development of offshore wind farms in the area would provide hard substrata for the settlement of sessile benthos. Since the seafloor of the Adriatic is predominantly sedimentary this may alter the larval connectivity of benthic populations in the region. Here, we simulated the release of larvae from benthic populations along the coasts of the Adriatic Sea using coupled bio-physical models and investigated the effect of pelagic larval duration on dispersal. Our model simulations show that currents typically carry particles from east to west across the Adriatic, whereas particles released along western coasts tend to remain there with the Puglia coast of Italy acting as a sink for larvae from benthic populations. We identify areas of high connectivity, as well as areas that are much more isolated, and discuss how these results can be used to inform marine spatial planning and the licensing of offshore marine renewable energy developments. [ABSTRACT FROM AUTHOR]

The interactive effect is significant in the Chinese stock market, exacerbating the abnormal market volatilities and risk contagion. Based on daily stock returns in the Shanghai Stock Exchange (SSE) A-shares, this paper divides the period between 2005 and 2018 into eight bull and bear market stages to investigate interactive patterns in the Chinese financial market. We employ the Least Absolute Shrinkage and Selection Operator (LASSO) method to construct the stock network, compare the heterogeneity of bull and bear markets, and further use the Map Equation method to analyse the evolution of modules in the SSE A-shares market. Empirical results show that (1) the connected effect is more significant in bear markets than bull markets and gives rise to abnormal volatilities in the stock market; (2) a system module can be found in the network during the first four stages, and the industry aggregation effect leads to module differentiation in the last four stages; (3) some stocks have leading effects on others throughout eight periods, and medium- and small-cap stocks with poor financial conditions are more likely to become risk sources, especially in bear markets. Our conclusions are beneficial to improving investment strategies and making regulatory policies. [ABSTRACT FROM AUTHOR]

A functional region is usually defined as a region characterized by not only interaction cohesiveness but also function diversity. The former indicates that places within a region should share a cohesive interacting pattern of human movements, such as commute or daily travel; the latter means that the places should possess diverse functions like residence, commerce, etc. Due to these two characteristics, people living and moving in one region can acquire almost all their daily needs, and they do not need to frequently move to other regions. In other words, a functional region resembles a life circle. Based on this definition, many previous studies have developed different methods to analyse human flow data so that they can properly profile different types of functional regions like Local labour market area (LLMA) or daily urban system (DUS), which helps people better understand how functional regions compose an area. Existing methods have considered interaction cohesiveness in a functional region, but function diversity is ignored. This is because these studies only used specific-purpose-trip data, such as journey to work or to shop. These kinds of trips describe people moving from their home to their office or to shopping malls, so each trip has an inherent distinction of functions between its origin and destination; thus, a functional region consisting of these trips indirectly owns function diversity. However, this framework has some problems. First, it cannot profile a functional region shaped by composite functions, for only one trip purpose can be taken into account each time. Second, collecting the data of specific-purpose trips needs either some questionnaires or some surveys, so it is often costly and time consuming. Finally, the purpose-unrecorded-trip data provided by rapidly developed smart card systems in several countries can be collected quickly, yet it cannot properly profile functional regions through the aforementioned framework due to the lack of trip purposes. Therefore, we propose a new spatial partitioning algorithm which can simultaneously consider both interaction cohesiveness and function diversity in profiling functional regions. Compared with previous methods, the proposed algorithm can better profile functional regions from many purpose-unrecorded-trip data based on the inclusion of land-use data. Our results provide deeper insights for understanding how various functions of lands and human movements together shape the boundary of a region. It could be significant reference for developing policies for urban planning or public transportation management. [ABSTRACT FROM AUTHOR]