"Form-Based Processing/Extraction Based on Templates"

thapli_64
thapli_64 New Altair Community Member
edited November 5 in Community Q&A

Hi All,

I was wondering if there is any precedent (in RapidMiner) of processing and extracting text information from forms where we have the templates, using either a rule-based approach or possibly even Machine Learning? 

Tagged:

Best Answer

Answers

  • sgenzer
    sgenzer
    Altair Employee

    hello @thapli_64 - so there are lots of ways to extract and process text from a variety of sources.  What format are your templates in?

     

    Scott

     

  • thapli_64
    thapli_64 New Altair Community Member

    @sgenzer thanks for your response. So I checked, and it seems we wont really have templates before hand per se. but we'll have lots of documents (forms) that have a particular layout. We'll be performing OCR on them, so we'll have TXT and XML files. So I guess those will serve as a starting point. I hope this clarifies things? 

  • sgenzer
    sgenzer
    Altair Employee

    hi @thapli_64 - yes that all makes sense and all very doable.  You're going to want to use the Text Processing extension for your TXT files and the Read XML operator for your XML files.  Then you are off and running.  :)

     

    Scott

  • thapli_64
    thapli_64 New Altair Community Member

    @sgenzer I've been trying to work with the Read XML operator and running into serious issues. Can you point me to some tutorial or or other resource that covers that operator so I can figure out if I'm working with it properly?

  • sgenzer
    sgenzer
    Altair Employee

    hmm that's a good question.  I actually don't know of a Read XML tutorial.  Perhaps you can post the xml file you're trying to read, along with your process XML, and I can take a look?

     

    Scott

     

  • thapli_64
    thapli_64 New Altair Community Member

    @sgenzer That's the trouble I've been having. I have been bery successful finding resources on almost every question I've had, but Read XML has ben particually vexing. Is there any place we can put in a request for this? Maybe something that could also be covered in an upcoming blog post, webinar or office hours?

     

    My company is an RM customer. Is there someone we can reach out to for one on one help with this?

     

    Unfortunately, the XML files contain sensitive data so I can't share them. I will try to see if I can create or procure dummy files with fake data.

  • sgenzer
    sgenzer
    Altair Employee
    Answer ✓
    hi @thapli_64 - wrote you a PM.

    Scott
  • thapli_64
    thapli_64 New Altair Community Member

    thanks Scott

Categories