I have some letters that I'm trying to run through Advanced Capture. The only place where the full name appears is within the addressee field at the top of the letter, like so:
John Doe 123 Main St City, ST, Zip
I've tried using the Person Name Extraction option but it keeps grabbing city/state as the name data. Does anyone have any tips on how I can accurately grab first name/last name from this type of data instance? They are two separate keyword fields on the doc type and they're the only keywords I need. Thanks in advance!
I agree with Jessica that creating your own regular expression and saving that regular expression to the library in Advanced Capture is going to be your best option.
Because you are looking to extract two keyword values, you will need to use Capture Groups in your regular expression and configure two overlapping keyword field zones on your Advanced Capture Template (Form). I think in order to explain what I mean, it would be easier to record a quick proof of concept for you. Here is a video that explains my idea.
If you have Premium Subscription, there are Hands-On Experience Labs that come with a Virtual Machine - it is this machine that I used to build my proof of concept. https://training.hyland.com/courses/P1805-Advanced-Capture-Hands-On-Lab This exact scenario that you described is not covered in the lab steps (which is why I recorded a video to explain what I am thinking) but you can always feel free to try out additional ideas in the Virtual Machine.
I had to set up something like this in one of our forms to capture company name. First make sure your zone encompasses the whole address and a little extra room around it for any scaling/translation that may occur. Next choose 'Find by tag/expression' instead of 'Person name extraction'. Then under Filtering select 'Find matching reg. expression', clicking F2 in the box opens up your library. I saved the below regular expression to my library.
^([ \u00c0-\u01ffa-zA-Z'])+$
This was able to find in my case a company name that was on the first line no matter how many words were in the name. It also didn't pick up any part of the address whether it started with numbers or PO BOX. I hope this can work for you as well.
I agree with Jessica that creating your own regular expression and saving that regular expression to the library in Advanced Capture is going to be your best option.
Because you are looking to extract two keyword values, you will need to use Capture Groups in your regular expression and configure two overlapping keyword field zones on your Advanced Capture Template (Form). I think in order to explain what I mean, it would be easier to record a quick proof of concept for you. Here is a video that explains my idea.
If you have Premium Subscription, there are Hands-On Experience Labs that come with a Virtual Machine - it is this machine that I used to build my proof of concept. https://training.hyland.com/courses/P1805-Advanced-Capture-Hands-On-Lab This exact scenario that you described is not covered in the lab steps (which is why I recorded a video to explain what I am thinking) but you can always feel free to try out additional ideas in the Virtual Machine.
