AWS neptune-analytics documentation change
Summary
Added documentation for Louvain algorithm and its mutate variant, including use cases like fraud detection.
Security assessment
The change introduces new algorithm documentation with mentions of security-related use cases (fraud detection), but does not address a specific security vulnerability or add explicit security feature documentation.
Diff
diff --git a/neptune-analytics/latest/userguide/clustering-algorithms.md b/neptune-analytics/latest/userguide/clustering-algorithms.md index f962a09f5..c7590f74e 100644 --- a//neptune-analytics/latest/userguide/clustering-algorithms.md +++ b//neptune-analytics/latest/userguide/clustering-algorithms.md @@ -26,0 +27,4 @@ Identifying weakly-conected components helps in understanding the overall connec + * [louvain](./louvain.html) – The Louvain algorithm is a hierarchical community detection method that identifies groups of densely connected nodes within networks. It works by optimizing modularity - a measure comparing internal community connection density against random networks with the same degree distribution. Through iterative local optimization and network aggregation, it efficiently detects community structures in large networks. The algorithm can run in both weighted and unweighted modes; when an edge weight property is specified, it operates in weighted mode. Louvain is valuable in social networks for identifying user communities, in biological networks to discover functional modules or protein complexes, and in financial networks to detect market segments or potential fraud rings. + + * [louvain.mutate](./louvain-mutate.html) – This Louvain variant stores the calculated community ID of each node as a property of the node, enabling persistent community assignments for applications like customer segmentation, network infrastructure optimization, and research community identification. +