Hello,
I would like to use FME to search for partial name matches (essentially a fuzzy string comparison) between two columns in different datasets.
The challenge I'm facing is that it’s not a straightforward ID-based comparison. I want to pick a name from a row in one dataset and search for the closest match in a column of names in another dataset.
Does anyone know a way to achieve this in FME, perhaps without relying on IDs?
Thank you!
If you have a lot of data I’d try to do this in the database, if possible. It’s going to be a lot faster than loading everything into FME and doing all the matching there.
E.g. using Postgres: https://www.postgresql.org/docs/current/fuzzystrmatch.html
If you have to stick with FME, maybe this custom transformer can help: https://hub.safe.com/publishers/safe-lab/transformers/fuzzystringcomparer
How much data are we talking about? What I have done is merging the searchstring to all target records (FeatureMerger 1=1) and use the FuzzyStringComparer from the FME hub to find the best match. But it you have a lot of search and a lot of target records, this can take a while. Attached a sample matching 100 search by 100 target values.