Group By/ Fuzzy Match

[danielmaguire]

Hi,

I have attached an excel below, which I am having trouble with. Essentially I have grouped the data by company name. However I am also trying to match column A and find the corresponding value in column B. However as you can see, column a and b differ quite a lot in terms of abbreviations, spellings etc.

If anyone could help provide a workflow which achieves the aforementioned end result that would be greatly appreciated.

Thanks,

Daniel

[BrandonB]

@danielmaguire normally I would take a fuzzy match approach for comparing two lists of business names, but this will definitely need tweaking given how close some non-identical company names are in your list. I have attached the fuzzy matching workflow as a good starting point, but if you run it and take a look at the results you will see where there are some matches that aren't desirable with the current settings. 

 

[mceleavey]

Hi @danielmaguire ,

I've built a fuzzy match workflow for you. 

The first step is to take out those that match 100%.

The remainder I've matched through fuzzy matching, but you'll need to use a bit of trial and error with the algorithm settings.

M.

[atcodedog05]

Was thinking through how can be done and this is amazing @BrandonB. Still little bit complicated.

Great exposure 🙂  

[BrandonB]

Here are a couple of training videos if you are interested in the topic:

https://community.alteryx.com/t5/Videos/Fuzzy-Matching-for-Beginners/td-p/330575

https://community.alteryx.com/t5/Videos/Fuzzy-Matching-Intermediate-Users/td-p/43852

[danielmaguire]

Hi @BrandonB and @mceleavey .

Thanks for providing those workflows. As you mentioned it is extremely difficult to get the match correct given the similarities between the two columns. However, I will continue to play around with the data to see if I can get it to work.

Thanks again,

Daniel

[mceleavey]

HI @danielmaguire ,

Yeah, once you have the workflow set up you can play around with it, but I would recommend stripping out components of the string you wish to match and attempting multi-fuzzy matches. Try stripping out the single letter (Type A, for example) then fuzzy matching the remaining string and matching the letters like for like.

Fuzzy matching is an art, not a science 🙂

M.