Figure 2: The merging and mapping step of the ETL process.
At the first step on the left of the figure, the data source’s sub-graphs are stacked together. The second step describes the copy of each data source node into the mapper graph with only their identifiers as attributes. In step three, similar nodes are merged together in the mapper graph and step 4 transfers the edges of the data source graphs into the mapper graph. The mapper nodes are connected to their original source nodes via an edge.