How the Network is calculated

The Network is a dynamic view that grows and changes with each new finding and linkage you create. In this article, we share a bit of insight into how the Network is calculated.

Linkages draw finding bubbles closer

The Network is based on a commonly used network layout algorithm called Force-Directed Graph. In a nutshell, the layout of the Network is determined by two types of connections between the “nodes” such as Signal clusters, Trends, and Megatrends.

Thick lines represent linkages between findings. Someone in your account has made these linkages, or they came with the data set.

Human-made linkages have the most effect. So, if you connect two trends with a linkage, they will have a force driving them closer.

Shared Signals also affect the layout

The other type of connection is formed if e.g. two Trends are connected to the same Signals and thus have a non-direct connection.

When FIBRES notices that two findings share a significant number of linked Signals, it connects them with a thin line on the Network view.

These implicit linkages have a weaker effect but still will impact the layout the same way as direct linkages driving those trends closer to each other.

Unlinked finding bubbles drift apart

‌Finally, each finding has a repelling force that stops them from getting too close or overlapping. Thus, if you have findings or groups of findings with no linkages between them, they will drift apart.

This will help you spot related topic clusters and, on the other hand, identify lone findings with no linkages yet to other content.