GraphAcademy

Graph Data Science

modularity

A score for how much more densely connected the nodes within a group are than they would be by chance.

Example

The same six people, grouped two ways.

Grouping each triangle on its own scores 0.36. Seven relationships join the six people, and only one of the seven crosses between the groups.

Putting all six in a single group scores 0.

Lessons that use this term

The lesson and course links below open in a new tab.

9 lessons use this term. Results are ordered by term density.

Graph Data Science in Practice26 mentions

  • Louvain Community DetectionCommunity Detection for FraudExplain how Louvain finds communities using modularity optimization Interpret modularity scores to assess community quality Configure Louvain for different use cases19 mentions
  • Reducing Search Space with LouvainCommunity Detection for FraudYou've learned how Louvain finds communities by optimizing modularity3 mentions
  • Applying AlgorithmsGDS FoundationsLouvain groups nodes by maximising "modularity"--connection density within groups2 mentions
  • Degree Centrality & WCCCommunity Detection for FraudAspect | Louvain | WCC | | --- | --- | --- | | Logic | Optimizes modularity (density) | Any path = same component | | Output | Dense clusters | All reachable nodes | | Determinism | Can vary2 mentions

Analyze Graph Data with Python4 mentions

  • Louvain Community DetectionGDS Python ClientPractice](/courses/workshop-gds/) workshop to detect fraud rings by optimizing modularity3 mentions
  • Scaled Properties and FastRP EmbeddingsGDS Python ClientInput | Graph structure only | Structure + content + centrality | | Method | Modularity optimization | Distance in embedding space | | Clusters | Found automatically | You specify k | | Best for | Finding1 mention

Get started with Graph Data Science4 mentions

  • Configuring projections with unweighted relationshipsWorking with algorithmsLower modularity: indicating lower-quality communities Higher community count: indicating greater difficulty in identifying1 mention
  • Understand the five execution modesWorking with algorithms: The final modularity score—a measure of how well-divided the network is into communities (higher is better, range -0.5 to 1.0)2 mentions
  • Algorithms overviewWorking with algorithmsLouvain and Leiden work by maximizing "modularity", or the density of connections within groups compared to the density between1 mention

Related terms

Louvain algorithm
A community detection algorithm that repeatedly merges nodes into groups for as long as merging raises modularity.9 shared lessons
node embedding
A list of numbers that stands in for a node's position in the graph, so that nodes in similar positions get similar lists.3 shared lessons
centrality
How important a node is within a graph. Each centrality algorithm defines importance differently.6 shared lessons
community detection
A family of algorithms that group nodes by how they connect. Each algorithm defines a community differently.7 shared lessons
weakly connected components
Groups of nodes in which every node is reachable from every other, once relationship direction is ignored. The GDS algorithm that finds them takes the same name.3 shared lessons
betweenness centrality
A score for each node equal to how often it lies on the shortest paths between other nodes.3 shared lessons
FastRP
A node embedding algorithm, short for Fast Random Projection. It builds each node's vector by combining random vectors drawn from the nodes around it.3 shared lessons
degree centrality
A score for each node equal to its number of outgoing relationships.5 shared lessons

All glossary terms