ISM and DEMATEL are similar to general network graph in that the adjacency matrix is the starting point for the analysis .
However, since it can be used as a method for analyzing the hierarchy of the system, it is a method to be introduced because it is System Engineering and Decision Making .
The adjacency matrix handled by ISM and DEMATEL represents a causal relationship. The orientation represented by the adjacency matrix is ??used to represent the relationship between cause and effect. Since it is a questionnaire, I will make this procession.
This is the first difference because general network graphs deal with things other than causality.
In a typical network graph, you create a network graph directly from the adjacency matrix. Adjacency matrices and network graphs look different, but they are essentially the same.
On the other hand, in ISM and DEMATEL, the adjacency matrix is ??processed before the network graph is created.
Suppose A causes C, B also causes C, and A also causes B.
In a typical network graph, A connects to B and C (left figure). On the other hand, ISM and DEMATEL do not connect from A to C. Since this can be done, ISM and DEMATEL analyze it as if there is a hierarchy in the order of C, B, A.
ISM uses a matrix of only 0s and 1s. DEMATEL can also analyze non-binary matrices such as 0, 1, 2, and 3.
I have a library called "ISM" in R, but I'm not trying it because I get an unclear error.
To do ISM, you need to calculate the reachable matrix, but how to find the reachable matrix in R can be found on the Reachability Matrix by R.
