Fuzzy Match Tool

[ahmedraza_umer]

I have a list of keywords in a file and I have another Excel file (Product List) with descriptions in it. I need to find those keywords in the descriptions. Once I match that, each of the keywords comes under a certain product category. And then I will calculate the respective weights. I cannot perform the first task. Let's say the description says OD:79mm, 22w, Steel Pipe... now I have a keyword Steel Pipe. How do I match that? Is the fuzzy match tool the right one for it?

[JosephSerpis · Accepted solution]

Hi @ahmedraza_umer I mocked up a workflow that shows an alternative approach using the find and replace tool. Let me know what you think?

[ahmedraza_umer]

Thank you. It worked! 

I also need to extract the OD size from the description which is in mm. How do I extract that? Like OD: 88mm is in the center of description so is there a way we can extract these?

Many thanks!

[JosephSerpis · Accepted solution]

Hi @ahmedraza_umer  you can use regex to extract that information you just need to identify the pattern or patterns to extract that information. I amended the workflow and mocked up an example of how to do this based on the example you gave. The regex looks for two word characters e.g. OD then it looks for : then two digits e.g.79 then two more word characters. Therefore it extracts OD:79MM for that example. 

[ahmedraza_umer]

Thank you so much for helping me out. 

If the digits are lets say 115.6 so do I put  (\w{2}:\d{4}\w{2}) and it will take upto 4 digits? Because some are 2 digits while others are 3 or 4 so how do I incorporate all of them?

Also what if there is space between colon and digits and then again before mm. Can there be one formula to use for all?

[JosephSerpis · Accepted solution]

Hi @ahmedraza_umer increasing the number to 4 won't work in that example as their is a period so that would break the pattern as you would need to say their is a period in the string. What you are talking about in your post is the different patterns in your data that you need identify and then write the appropriate regex syntax for. I have amended my workflow to tackle some of patterns you identified and I've broken down it down so you can understand what the regex is doing. I would highly recommend the Parsing Data interactive lessons if you not done yet as it involves a numbers lessons about regex.