Help Center> Graph Engine Service> User Guide> Algorithms> Betweenness Centrality
Updated on 2024-01-03 GMT+08:00
View PDF

Betweenness Centrality

Overview

Betweenness centrality is a measure of centrality in a graph based on shortest paths. This algorithm calculates shortest paths that pass through a vertex.

Application Scenarios

The Betweenness Centrality algorithm can be used for tracing man-in-the-middle in social networks and risk control networks and identifying key vertices in transportation networks. This algorithm is widely used for social networking, financial risk control, transportation networking, and city planning.

Parameter Description

Table 1 Algorithm parameters

Parameter

Mandatory

Description

Type

Value Range

Default Value

directed

No

Whether an edge is directed

Boolean

The value can be true or false.

true

weight

No

Weight of an edge

String

The value can be an empty string. If this parameter is left blank, the weight and distance of this edge are 1 by default. You can set this parameter to a property of the edge, and the property value will be the weight. If the edge does not have the specified property, the weight is 1 by default.

NOTE:

The weight of an edge must be greater than 0.

-

seeds

No

Vertex ID

String

If the graph is large, betweenness calculation can be slow. You can set seeds to the sampling nodes for approximate calculation. The more seeds nodes, the closer results to the accurate calculation. The number of vertices cannot be greater than 100,000.

-

k

No

Number of samples

Integer

If the graph is large, betweenness calculation can be slow. You can set k to randomly select k sampling vertices from the graph. The larger value, the closer results to the accurate calculation. The value cannot be greater than 100,000.

-

When you perform approximate betweenness calculation, either seeds or k must be specified. If both are specified, seeds vertices will be sampled by default and k will be ignored.

Precautions

None

Example

Set weight="length", directed=true, seeds ="Lee,Alice" and view the result.

Parent topic: Algorithms