Dear Dr. Naik,

this is a really interesting question. Maybe the following papers, blogs and websites contribute to answering this question. The first paper is also available on researchgate:

Article A Generalizable NLP Framework for Fast Development of Patter...

In terms of software / programming languages, there are links to a Java implementation and a Matlab toolbox on the following website:

https://www.cs.ubc.ca/~murphyk/Software/CRF/crf.html

Alternatively, there is a C++ written toolkit termed CRF++ on the following website:

https://taku910.github.io/crfpp/

In addition, there are applications written in python programming language:

http://www.albertauyeung.com/post/python-sequence-labelling-with-crf/

The following blog on extracting data from recipes is also interesting:

https://open.blogs.nytimes.com/2015/04/09/extracting-structured-data-from-recipes-using-conditional-random-fields/

In addition, there is the following paper on domain knowledge for natural language understanding:

http://ceur-ws.org/Vol-1128/paper9.pdf

Yours sincerely,

Rainer