Joining on Similar names

[prpatel]

Hello Community,

I'm looking for some thoughts on the following situation.

I have two datafiles that have different data.  One is a Purchase request from, the other is a Purchase order report.  There are a few key columns that are common to both that I wish to use as a pam to ensure that I do the mapping correctly.  One of those columns is the "Name_of_Requester" column.  But there is a problem....in the purchase request from, the requester uses his/her prefered name ("Jenn Smith" or "Jim West"), but in the Purshase order report, their names are their offical names ("Jennifer Smith" or "James West").  

Is there a way in Alteryx that can detect these similar names?

Thanks in advance.

-Tesh.

[Matthew]

is your list of requesters a manageable size?

if so, you could just build a mapping table

[prpatel]

@Matthew ...this is exactly my thinking.  Can I build the mapping table within the workflow or does this need to be an excel sheet that I have to drag in all the time?  

I was hoping to avoid this step and use some sort of Fuzzy Logic match to allow me to do this, but your solution works, but requires additional maintenance to keep up the map.

Thanks though for responding.

-Tesh.

[Matthew]

have a look at the text input tool, it will allow you to manage the mapping table inside your workflow, not as a separate excel sheet

"fuzzy logic" is a bit of a vague term.. we could come up with some logic, but there will always be room for errors.. for example,we could try matching on the first initial and the last name, but what if you have a James Smith and a Jeremy Smith?

also, if you choose to go with this solution, would you mind clicking the "mark as solution" button? ;)

[FrederikE]

Is the Fuzzy Match Tool not reliable enough for you? That should be able to pick up most/all of these Name-Nickname combinations. 

[prpatel · Accepted solution]

@FrederikE...thanks for your input.  I'll have to play with it and see how it works out.

I was able to fix both reports to bring in company e-mails...so I'll use those and avoid the risks of missing records due to mis-match in names.