Comparing fields for similar data and group together

[RohanSheth]

I am trying to group together fields based on the similarity of data. Please see below for an example of what I'm trying to accomplish:

DATA:

Transaction	Employee1	Employee2	Employee3	Employee4	Employee5
CC1	X			X	X
CC2	X				X
CC3		X	X
CC4		X	X
CC5	X			X	X
CC6		X	X
CC7		X
CC8		X	X
CC9	X			X
CC10	X			X	X
CC11	X	X	X	X	X
CC12	X	X	X	X	X

My goal is for the workflow to look through the transactions that the employees are assigned to and group them together if they are similar by a certain parameter (hopefully a percentage). So for the above sample data, Employee1, 4, and 5 are pretty similar. So I would like the workflow to return these Employee fields and group them together. Similarly, Employee 2 and 3 are a close match so these should be a separate group if possible.

The data that I am ultimately trying to create groups with is much larger than the sample above, so it would be nice to have a tool setup to group these automatically. I've looked into using the Fuzzy Match and Make Group tools in conjunction, but I am unsure if the scope of the tools fit what I'm trying to do. If it is, how would I configure the tools based on the sample above?

Any insight would be greatly appreciated!

[neilgallen · Accepted solution]

You can use the fuzzy match and make group tools to do what you want. Ultimately you'd want a transaction string for each employee, and compare the strings to see how similar they are. The attached workflow is a start. You'd have to adjust the match percentages in order to meet your needs.

Fuzzy matching is an art, not science.

[RohanSheth]

Ok now I get it. So I needed to format it correctly by concatenating and transposing for the fuzzy joins to work. From the output, I can definitely group them together based on my preferences.

Definitely an art that would be neat to master 🙂 Thank you!