-
A method to compute the communicability of nodes through causal paths in temporal networks
Authors:
Agostino Funel
Abstract:
We present a method aimed to compute the communicability (broadcast and receive) of nodes through causal paths in temporal networks. The method considers all possible combinations of chronologically ordered products of adjacency matrices of the network snapshots and by means of a dam** procedure favors the paths that have high communication efficiency. We apply the method to four real-world netw…
▽ More
We present a method aimed to compute the communicability (broadcast and receive) of nodes through causal paths in temporal networks. The method considers all possible combinations of chronologically ordered products of adjacency matrices of the network snapshots and by means of a dam** procedure favors the paths that have high communication efficiency. We apply the method to four real-world networks of face-to-face human contacts and identify the nodes with high communicability. The accuracy of the method is proved by studying the spread of an epidemic in the networks using the susceptible-infected-recovered model. We show that if a node with high broadcast is chosen as the origin of the outbreak of infection then the epidemic spreads early while it is delayed and inhibited if the origin of infection is a node with low broadcast. Receiving nodes can be treated as broadcasters if the arrow of time is reversed.
△ Less
Submitted 5 August, 2021;
originally announced August 2021.
-
Causal Paths in Temporal Networks of Face-to-Face Human Interactions
Authors:
Agostino Funel
Abstract:
In a temporal network causal paths are characterized by the fact that links from a source to a target must respect the chronological order. In this article we study the causal paths structure in temporal networks of human face to face interactions in different social contexts. In a static network paths are transitive i.e. the existence of a link from $a$ to $b$ and from $b$ to $c$ implies the exis…
▽ More
In a temporal network causal paths are characterized by the fact that links from a source to a target must respect the chronological order. In this article we study the causal paths structure in temporal networks of human face to face interactions in different social contexts. In a static network paths are transitive i.e. the existence of a link from $a$ to $b$ and from $b$ to $c$ implies the existence of a path from $a$ to $c$ via $b$. In a temporal network the chronological constraint introduces time correlations that affects transitivity. A probabilistic model based on higher order Markov chains shows that correlations that can invalidate transitivity are present only when the time gap between consecutive events is larger than the average value and are negligible below such a value. The comparison between the densities of the temporal and static accessibility matrices shows that the static representation can be used with good approximation. Moreover, we quantify the extent of the causally connected region of the networks over time.
△ Less
Submitted 24 July, 2020; v1 submitted 7 May, 2020;
originally announced May 2020.
-
The Graph Structure of the Internet at the Autonomous Systems Level during Ten Years
Authors:
Agostino Funel
Abstract:
We study how the graph structure of the Internet at the Autonomous Systems (AS) level evolved during a decade. For each year of the period 2008-2017 we consider a snapshot of the AS graph and examine how many features related to structure, connectivity and centrality changed over time. The analysis of these metrics provides topological and data traffic information and allows to clarify some assump…
▽ More
We study how the graph structure of the Internet at the Autonomous Systems (AS) level evolved during a decade. For each year of the period 2008-2017 we consider a snapshot of the AS graph and examine how many features related to structure, connectivity and centrality changed over time. The analysis of these metrics provides topological and data traffic information and allows to clarify some assumptions about the models concerning the evolution of the Internet graph structure. We find that the size of the Internet roughly doubled. The overall trend of the average connectivity is an increase over time, while that of the shortest path length is a decrease over time. The internal core of the Internet is composed of a small fraction of big AS and is more stable and connected that external cores. A hierarchical organization emerges where a small fraction of big hubs are connected to many regions with high internal cohesiveness, poorly connected among them and containing AS with low and medium number of links. Centrality measurements indicate that the average number of shortest paths crossing an AS or containing a link between two of them decreased over time.
△ Less
Submitted 26 August, 2019; v1 submitted 13 February, 2019;
originally announced February 2019.
-
Analysis of the Web Graph Aggregated by Host and Pay-Level Domain
Authors:
Agostino Funel
Abstract:
In this paper the web is analyzed as a graph aggregated by host and pay-level domain (PLD). The web graph datasets, publicly available, have been released by the Common Crawl Foundation and are based on a web crawl performed during the period May-June-July 2017. The host graph has $\sim$1.3 billion nodes and $\sim$5.3 billion arcs. The PLD graph has $\sim$91 million nodes and $\sim$1.1 billion arc…
▽ More
In this paper the web is analyzed as a graph aggregated by host and pay-level domain (PLD). The web graph datasets, publicly available, have been released by the Common Crawl Foundation and are based on a web crawl performed during the period May-June-July 2017. The host graph has $\sim$1.3 billion nodes and $\sim$5.3 billion arcs. The PLD graph has $\sim$91 million nodes and $\sim$1.1 billion arcs. We study the distributions of degree and sizes of strongly/weakly connected components (SCC/WCC) focusing on power laws detection using statistical methods. The statistical plausibility of the power law model is compared with that of several alternative distributions. While there is no evidence of power law tails on host level, they emerge on PLD aggregation for indegree, SCC and WCC size distributions. Finally, we analyze distance-related features by studying the cumulative distributions of the shortest path lengths, and give an estimation of the diameters of the graphs.
△ Less
Submitted 6 March, 2018; v1 submitted 15 February, 2018;
originally announced February 2018.
