0
$\begingroup$

There are a range of models, articles and information available which extract keywords from unstructured documents. Examples are Spacy, POS tagging, NER, entity extraction etc

What I am looking for is key value pair extraction and not able to find much relevant articles or models or even datasets. Few examples of key value pairs are:
Name : Sandeep
Company Name: Sandeep
Total Amount $10000
Invoice INV110011
or INV | INV110011

There can be a range of formats in which key value pairs can come. In above examples, person name and company name is same, and without much context models like spacy etc get confused.
Can someone please point to some models or information or articles or datasets which are trying to extract this information.

$\endgroup$
1
  • $\begingroup$ This is mostly structured data, so imho it's usually faster and easier to simply design regular expressions for all the possible formats. If you want to train a NER model, the data should be annotated simply as two categories: key and value. Then the model should be able catch the syntax, but it's a bit of an overkill. $\endgroup$
    – Erwan
    Apr 27 at 13:29

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.