Skip to content

Welcome to the Collaboration Pattern Detection Framework documentation!

Tip

This project and documentation is under active development.

This Python3-based framework conducts Collaboration Pattern Detection using methods from Process Mining, Frequent Subgraph Mining & Graph Set Clustering techniques. Resulting (process) graph sets / patterns are stored in a Graph Collection and can be exported as a file, or uploaded/downloaded from or to a Neo4j graph database.

This framework is used in the Social Process Mining project to detect Collaboration Pattern inside Enterprise Collaboration Systems (ECS).

Overview

  1. Preprocessing: This package provides an API for preprocessing the event logs and prepare them for the collaboration pattern detection discovery algorithm.
  2. Process Mining: This package provides an API for mining graphs (process discovery) from event logs using Process Mining discovery algorithms.
  3. Frequent Subgraph Mining: The generated graphs are used as input for Frequent Subgraph Mining.
  4. Clustering Graph Sets: The resulting sub-graphs can be clustered using graph set clustering.