Fuzzy Matching Question

[FJ_Shafeek]

Hello,

I am trying to fuzzy match using the merge option across two datasets. As you can see from this screenshot of the data, I would expect the colored rows to match each other:

Here is my configuration of the tool itself:

With each custom match style set to the following:

I know it is something to do with the 'Specific Method' column since the other two columns match perfectly, and I know that the extra words are throwing the match off but I feel like there should be some configuration that matches them. I'm attaching the workflow as well, and any help is massively appreciated!

[BenMoss]

Hi @FJ_Shafeek I'd start looking at the way in which you are generating keys and also consider whether you wish to lay over match function information, perhaps using the 'company name' as a base for your match styles.

In your example, if you simply change the generate keys to 'double metaphone' and the maximum key length to 4, then you begin to get matches between objects.

One thing I will say is in your sample data, you say you expect 'these rows' to match, but there are other rows which are identical (in terms of matching fields) to those you have highlighted which you haven't, for example, the match fields you are using are the same for record 1361 (Which you haven't highlited), and 1366 which you have matched.

Ben