Jaccard Index

Similarity measures are used to determine how many features two entities have in common.

The Jaccard Index is the ratio of common neighbours between two nodes compared to the total number of neighbours of both nodes. The more neighbours two nodes have in common, the more similar they are. The features used to determine similarity with the Jaccard Index are a node's neighbours.

Parameters

• Transaction Types - Calculate on only the subgraph of transactions of these types
• Include Incoming - Include incoming connections
• Include Outgoing - Include outgoing connections
• Include Undirected - Treat undirected connections as bidirectional connections
• Minimum Common Features - Only calculate similarity between nodes that share at least this many features
• Selected Only - Calculate using only selected graph elements