This is an early prototype for the ADIPEC paper "Applying Machine Learning NLP Algorithm for Reconciliation Geology and Petrophysics in Rock Typing"
Please check the paper in this repo 📄spe-216223-ms-2.pdf or here: Applying Machine Learning NLP Algorithm for Reconciliation Geology and Petrophysics in Rock Typing
Problem: almost impossible to use thin section description for rock typing. Lots of plain text of every sample needed to be read.
Aim: improve rock typing to improve "Porosity-Permeability"
Solution: convert the text using pattern recognition, plain text => squared logs
Tools: large language model (LLM) -- Open AI API (model = 'gpt-3.5-turbo')
How: parameter = noun, value of the parameter = adjective describing the noun
