The Internet Archive has a preservation copy of this work in our general collections.
The file type is
Community finding algorithms for networks have recently been extended to dynamic data. Most of these recent methods aim at exhibiting community partitions from successive graph snapshots and thereafter connecting or smoothing these partitions using clever time-dependent features and sampling techniques. These approaches are nonetheless achieving longitudinal rather than dynamic community detection. We assume that communities are fundamentally defined by the repetition of interactions among aarXiv:1111.2018v1 fatcat:e637qdgogrd2tgd45id76fswxa
more »... of nodes over time. According to this definition, analyzing the data by considering successive snapshots induces a significant loss of information: we suggest that it blurs essentially dynamic phenomena - such as communities based on repeated inter-temporal interactions, nodes switching from a community to another across time, or the possibility that a community survives while its members are being integrally replaced over a longer time period. We propose a formalism which aims at tackling this issue in the context of time-directed datasets (such as citation networks), and present several illustrations on both empirical and synthetic dynamic networks. We eventually introduce intrinsically dynamic metrics to qualify temporal community structure and emphasize their possible role as an estimator of the quality of the community detection - taking into account the fact that various empirical contexts may call for distinct 'community' definitions and detection criteria.
, Danisch and Tabourier 2021 See https://perf.wiki.kernel.org. ... Cache-mr: proportion of data that was not found in cache (L1, https://github.com/lecfab/rescience-gorder ReScience C 7.1 (#3) -Lécuyer, Danisch and Tabourier 2021 ReScience C 7.1 (#3) -Lécuyer ...doi:10.5281/zenodo.4836230 fatcat:ca57qknx5fg3npchxqduwkiufe
Community finding algorithms for networks have recently been extended to dynamic data. Most of these recent methods aim at exhibiting community partitions from successive graph snapshots and thereafter connecting or smoothing these partitions using clever time-dependent features and sampling techniques. These approaches are nonetheless achieving longitudinal rather than dynamic community detection. We assume that communities are fundamentally defined by the repetition of interactions among adoi:10.1016/j.comnet.2011.10.024 fatcat:7po32cup3fbqnnqlxys4ck7sq4
more »... of nodes over time. According to this definition, analyzing the data by considering successive snapshots induces a significant loss of information: we suggest that it blurs essentially dynamic phenomena -such as communities based on repeated inter-temporal interactions, nodes switching from a community to another across time, or the possibility that a community survives while its members are being integrally replaced over a longer time period. We propose a formalism which aims at tackling this issue in the context of time-directed datasets (such as citation networks), and present several illustrations of both empirical and synthetic dynamic networks. We eventually introduce intrinsically dynamic metrics to qualify temporal community structure and emphasize their possible role as an estimator of the quality of the community detection -taking into account the fact that various empirical contexts may call for distinct 'community' definitions and detection criteria.
Link prediction appears as a central problem of network science, as it calls for unfolding the mechanisms that govern the micro-dynamics of the network. In this work, we are interested in ego-networks, that is the mere information of interactions of a node to its neighbors, in the context of social relationships. As the structural information is very poor, we rely on another source of information to predict links among egos' neighbors: the timing of interactions. We define several features toarXiv:1512.04776v1 fatcat:f6lynqiunffo7n5rsorzwzrsze
more »... pture different kinds of temporal information and apply machine learning methods to combine these various features and improve the quality of the prediction. We demonstrate the efficiency of this temporal approach on a cellphone interaction dataset, pointing out features which prove themselves to perform well in this context, in particular the temporal profile of interactions and elapsed time between contacts.
SpringerBriefs in Statistics
Analyzing interactions over time plays a key role in many contexts: recommender systems (who buys which product and when), contacts between individuals (message exchanges, physical proximity or phone calls, for instance), and transaction analysis (like money or data transfers) are typical examples. As a consequence, much effort is devoted to the analysis of such data with approaches like temporal networks, time-varying graphs or link streams [4, 2, 6] . Predicting future interactions is adoi:10.1007/978-3-030-14683-2_6 fatcat:ljqkjhcwjnd6herze67kuskhu4
more »... l question in all these contexts, but the problem is traditionally addressed by merging interactions into a graph or series of graphs, called snapshots [7, 9, 12] . This has the key advantage of building a bridge with the powerful formalism and tools of graph theory, but at the cost of important information losses. More importantly, we argue that this approach misses interesting variants of the problem itself. The goal of this chapter is to deepen our understanding of these interaction prediction problems. To do so, we formalize them within the link stream framework, which makes it possible to fully capture both the temporal and structural nature of data. This leads to several meaningful problem definitions, that raise quite different challenges, as well as relations between them and classical approaches. We focus here on problem definitions and comparisons; resolving some of them has already received attention [5, 3, 1] but unifying them into the same framework leads to a better understanding of the whole and the identification of new variants of interest. We also show that this helps to identify general approaches to tackle them. Throughout this chapter, we assume a standard approach for solving prediction problems. First, one designs a model in order to make a prediction based on the fundamental assumption that future behaviors can be predicted from past observations.
Listing triangles is a fundamental graph problem with many applications, and large graphs require fast algorithms. Vertex ordering allows to orient the edges from lower to higher vertex indices, and state-of-the-art triangle listing algorithms use this to accelerate their execution and to bound their time complexity. Yet, only two basic orderings have been tested. In this paper, we show that studying the precise cost of algorithms instead of their bounded complexity leads to faster solutions.arXiv:2203.04774v1 fatcat:46hmf653nvah7edcm4y3upzhoy
more »... introduce cost functions that link ordering properties with the running time of a given algorithm. We prove that their minimization is NP-hard and propose heuristics to obtain new orderings with different trade-offs between cost reduction and ordering time. Using datasets with up to two billion edges, we show that our heuristics accelerate the listing of triangles by an average of 30% when the ordering is already given as an input, and 15% when the ordering time is included.
In this paper, we propose data clustering techniques to predict temporal characteristics of data consumption behavior of different mobile applications via wireless communications. While most of the research on mobile data analytics focuses on the analysis of call data records and mobility traces, our analysis concentrates on mobile application usages, to characterize them and predict their behavior. We exploit mobile application usage logs provided by a Wi-Fi local area network service providerdoi:10.1016/j.comcom.2016.04.026 fatcat:bcsx3klti5f6pgrs6d4q5s2nwm
more »... to characterize temporal behavior of mobile applications. More specifically, we generate daily profiles of "what" types of mobile applications users access and "when" users access them. From these profiles, we create usage classes of mobile applications via aggregation of similar profiles depending on data consumption rate, using three clustering techniques that we compare. Furthermore, we show that we can utilize these classes to analyze and predict future usages of each mobile application through progressive comparison using distance and similarity comparison techniques. Finally, we also detect and exploit outlying behavior in application usage profiles and discuss methods to efficiently predict them.
EPJ Data Science
k = 4 +19.3% k = 5 +21.4% k = 6 +22.3% k = 7 +22.5% k = 8 +25.5% k = 9 +25.5% k = 10 +28.1% k = 11 +30.9% k = 12 +26.4% k = 13 +33.1% k = 14 +36.2% k ≥ 15 +51.6% © 2016 Tabourier ...doi:10.1140/epjds/s13688-015-0062-0 fatcat:xgxcbojcpvglrdfd4twhsdpjh4
The generation of random graphs using edge swaps provides a reliable method to draw uniformly random samples of sets of graphs respecting some simple constraints, e.g. degree distributions. However, in general, it is not necessarily possible to access all graphs obeying some given con- straints through a classical switching procedure calling on pairs of edges. We therefore propose to get round this issue by generalizing this classical approach through the use of higher-order edge switches. ThisarXiv:1012.3023v2 fatcat:pfwvxuzaovb67j4qgsudkj2uaa
more »... method, which we denote by "k-edge switching", makes it possible to progres- sively improve the covered portion of a set of constrained graphs, thereby providing an increasing, asymptotically certain confidence on the statistical representativeness of the obtained sample.
The generation of random graphs using edge swaps provides a reliable method to draw uniformly random samples of sets of graphs respecting some simple constraints, e.g. degree distributions. However, in general, it is not necessarily possible to access all graphs obeying some given constraints through a classical switching procedure calling on pairs of edges. We therefore propose to get round this issue by generalizing this classical approach through the use of higher-order edge switches. Thisdoi:10.1145/1963190.2063515 fatcat:xfyjffistjf2zb4aw7z5xv4vly
more »... thod, which we denote by "k-edge switching", makes it possible to progressively improve the covered portion of a set of constrained graphs, thereby providing an increasing, asymptotically certain confidence on the statistical representativeness of the obtained sample.
Citation cascades in blog networks are often considered as traces of information spreading on this social medium. In this work, we question this point of view using both a structural and semantic analysis of five months activity of the most representative blogs of the french-speaking community.Statistical measures reveal that our dataset shares many features with those that can be found in the literature, suggesting the existence of an identical underlying process. However, a closer analysis ofarXiv:1306.0424v1 fatcat:2fnagfin3napnomnx5h7qcfwy4
more »... the post content indicates that the popular epidemic-like descriptions of cascades are misleading in this context.A basic model, taking only into account the behavior of bloggers and their restricted social network, accounts for several important statistical features of the data.These arguments support the idea that citations primary goal may not be information spreading on the blogosphere.
Capturing both the structural and temporal aspects of interactions is crucial for many real world datasets like contact between individuals. Using the link stream formalism to capture the dynamic of the systems, we tackle the issue of activity prediction in link streams, that is to say predicting the number of links occurring during a given period of time and we present a protocol that takes advantage of the temporal and structural information contained in the link stream. Using a supervisedarXiv:1804.01465v2 fatcat:dcadbot2uzchpbbbavsqzsrml4
more »... rning method, we are able to model the dynamic of our system to improve the prediction. We investigate the behavior of our algorithm and crucial elements affecting the prediction. By introducing different categories of pair of nodes, we are able to improve the quality as well as increase the diversity of our prediction.
Cet article étudie la robustesse d'une économie régionale à un choc exogène tel qu'une catastrophe naturelle. Il est basé sur une modèle dynamique qui représente une économie régionale comme un réseau d'unités de production constitué à partir d'une table entrée sortie sectorielle. Les résultats suggèrent que les pertes de production liées aux catastrophes naturelles dépendent de l'hétérogénéité des pertes directes et de la structure du réseau économique. Deux indices agrégés la concentration etdoi:10.1016/j.jedc.2011.10.001 fatcat:qcod2u7ofngt5j6vpxikinmbgy
more »... le regroupement apparaissent comme déterminants dans la résilience à un choc, orant des possibilités de stratégies d'amélioration de la résilience. Abstract This article proposes a theoretical framework to investigate economic robustness to exogenous shocks such as natural disasters. It is based on a dynamic model that represents a regional economy as a network of production units through the disaggregation of sectorscale Input-Output tables. Results suggest that disaster-related output losses depend on direct losses heterogeneity and on the economic network structure. Two aggregate indexes concentration and clustering appear as important drivers of economic robustness, oering opportunities for robustness-enhancing strategies. Modern industrial organization seems to reduce short-term robustness in a trade-o against higher eciency in normal times.
We discuss how spreading processes on temporal networks are impacted by the shape of their inter-event time distributions. Through simple mathematical arguments and toy examples, we find that the key factor is the ordering in which events take place, a property that tends to be affected by the bulk of the distributions and not only by their tail, as usually considered in the literature. We show that a detailed modeling of the temporal patterns observed in complex networks can changedoi:10.1140/epjb/e2013-40456-9 fatcat:y4q27nlxbvh33ojfddyexuuyoy
more »... the properties of a spreading process, such as the ergodicity of a random walk process or the persistence of an epidemic.
Without having direct access to the information that is being exchanged, traces of information flow can be obtained by looking at temporal sequences of user interactions. These sequences can be represented as causality trees whose statistics result from a complex interplay between the topology of the underlying (social) network and the time correlations among the communications. Here, we study causality trees in mobile-phone data, which can be represented as a dynamical directed network. Thisdoi:10.1371/journal.pone.0028860 pmid:22216128 pmcid:PMC3247222 fatcat:5wcafmyxwfbvjazpdqa6a5zc3i
more »... presentation of the data reveals the existence of super-spreaders and super-receivers. We show that the tree statistics, respectively the information spreading process, are extremely sensitive to the in-out degree correlation exhibited by the users. We also learn that a given information, e.g., a rumor, would require users to retransmit it for more than 30 hours in order to cover a macroscopic fraction of the system. Our analysis indicates that topological node-node correlations of the underlying social network, while allowing the existence of information loops, they also promote information spreading. Temporal correlations, and therefore causality effects, are only visible as local phenomena and during short time scales. These results are obtained through a combination of theory and data analysis techniques.
« Previous Showing results 1 — 15 out of 53 results