Centrality Metrics for Graphs

A graph can be seen as a visual representation of the relationships between various nodes. These nodes, which are also referred to as entities, depict the connected individuals or elements. Graph analytics is a term used to describe the process of extracting meaningful insights from a graph by studying the interactions between nodes connected by edges. This paper will focus on one such concept, centrality, and its importance in gaining a comprehensive understanding of the graph.

The Meaning of “Central”

Identifying important nodes in a network is a critical part of graph analytics, and the concept of centrality is essential to this process. Centrality seeks to identify the most significant or central nodes in a network, thus allowing them to be highlighted.

In order to properly assess the importance of a given node, it is essential to analyse multiple metrics. Centrality provides the essential analytical data regarding the node and the graph, allowing for a conclusion to be drawn from the observations.

Centrality measurements are a set of metrics that enable us to gain an insight into the multiple perspectives of a given node. These metrics enable us to extract data from a network, and thus gain an in-depth understanding of the network. Furthermore, for the purpose of graph visualisation, it is essential to comprehend the various measures available, as they all assign a distinct meaning to the significance of a node.

Gradient of importance

Here’s a quick review on the degree of a node in a graph before we get into degree centrality.

It is possible to construct both directed and undirected graphs. In an undirected graph, the degree of a node is equal to the number of edges connecting it to other nodes. In a directed graph, the degree is further partitioned into in-degree and out-degree. The in-degree of a node is the number of linkages occurring on the node, whereas the out-degree is the number of ties emanating from the node and connecting to other nodes.

By utilising degree centrality, the importance of a node in a network can be accurately quantified. This method of analysis is based on the premise that the significance of a node increases in direct proportion to its degree, or the number of connections it has. This can be used to identify influential individuals in a network, such as those with the most extensive networks of contacts, those who rapidly move between peers, and those who have access to the most pertinent information.

Useful in the study of financial data, account activity, etc., it is also one of the simplest centrality measurements of node connection.

Proximity to the centre

By analysing the degree to which a node is interconnected with all other nodes in the network, closeness centrality assigns a node’s importance. This is achieved by calculating the geodesic distance (GD), which is determined by the number of edges necessary to connect one node to another.

Sum up the GD between a node and every other node in the network for that node to get its closeness or GD.

Closeness Centrality can be utilised to pinpoint people who have the potential to quickly influence the entire network. Additionally, it can be employed to detect people or organisations that are isolated yet influential. Individuals who are of paramount importance to the organisation have the ability to access and sway crucial data. Furthermore, a graph-based keyphrase extraction approach can be used to anticipate the importance of words in a particular text.

Harmonic equivalence centre

More specifically, this centrality is a measure of closeness; it quantifies global reachability (GD) across links. It should be noted that when some of the nodes are outside the range of reachability, the harmonic centrality measurements provide a more accurate estimation of proximity.

Location in the centre of things

The betweenness centrality of an edge measures the extent to which it serves as an intermediary in pathways that connect other edges. This metric is used to determine how close a given node is to the shortest route through a network, as well as the proportion of that route that it occupies.

If a node has a high betweenness centrality, it means that it plays a disproportionately influential role in the network in terms of its impact on the other nodes. This metric provides useful insight into which links in the network are critical for the system to function effectively, and as such, has become an essential tool for maintaining system stability.

By utilising betweenness centrality, anti-terrorism authorities are able to analyse terrorist networks on a global scale. The data collected by these safeguards is then used to identify and neutralise potential threats, as well as monitor data transmission rates in telecommunications networks and track package deliveries over the internet. Furthermore, microbloggers can capitalise on this hub status to grow their Twitter following with the help of a recommendation engine. This enables them to determine which contacts should be made in order to grow their sphere of influence.

Centricity of Eigenvectors

The importance of a node can be measured using its eigenvector centrality, which considers the influence of its neighbouring nodes. To illustrate, imagine a node in a network. Each of its linked nodes should be evaluated to determine its eigenvector centrality. A high score indicates that the node is connected to and surrounded by influential nodes, thus making it a key component of the network.

When computing a node’s eigenvector centrality score, the connections to other nodes with higher scores are given more importance than those with lower scores. This means that the score of the node is determined by the amount of influence it has on other nodes, which is determined by the number and quality of the connections it has with other nodes.

By assessing the centrality of a node, it is possible to identify those nodes which have a wide-reaching influence on the entire network, rather than simply their immediate neighbours. Nodes which have a high centrality score will often have other nodes close by which also possess a high score. This measure of centrality can be used for a range of applications, such as Google’s PageRank algorithm, and for the analysis of virus and social networks.


PageRank is a form of eigenvector centrality, particularly well suited for directed graphs. It is calculated based on the effect of nodes on the directed graph, allowing it to measure the influence of the nodes in the specified direction.

Due to the fact that eigenvector centrality is most effective when applied to undirected graphs, no viable option was available for directed graphs until the emergence of PageRank. Popular social media platform, Twitter, is one of the many applications that leverage this prominence metric to suggest additional accounts to users that may be of interest.

In the healthcare and insurance industries, PageRank algorithms are utilised to identify weaknesses in the fraud detection system. Through a series of iterations that traverse a network that incorporates intersections and other points of contact, the system can even predict traffic flow in public places and on roads.


ArticleRank is a subset of PageRank that quantifies the influence of a node within a network by measuring its ability to affect other nodes through transitive citation links. Although this may appear to run contrary to a key principle of network theory, which suggests that relationships with nodes with a low number of outgoing connections are more significant than those with a high number, both PageRank and ArticleRank provide useful methods for obtaining meaningful results when analysing a network with a large number of nodes. With the knowledge of how to calculate the different kinds of centrality, it is possible to assess the nodes in a network from multiple perspectives, allowing for more accurate and robust conclusions.

Join the Top 1% of Remote Developers and Designers

Works connects the top 1% of remote developers and designers with the leading brands and startups around the world. We focus on sophisticated, challenging tier-one projects which require highly skilled talent and problem solvers.
seasoned project manager reviewing remote software engineer's progress on software development project, hired from Works blog.join_marketplace.your_wayexperienced remote UI / UX designer working remotely at home while working on UI / UX & product design projects on Works blog.join_marketplace.freelance_jobs