Understanding geological reports based on knowledge graphs using a deep learning approach

Description

This paper proposes a method for understanding geological report content using deep learning . GeoERE-Net is a triplet extraction network proposed for text knowledge mining in the geological field. This paper has been accepted by 《Computers & Geosciences》. If you refer to this repository, please cite it. The link of the paper: GeoERE-Net

Requirements

• pytorch 1.7
• pytorch_pretrained_bert
• numpy
• einops
• tqdm

Dataset

The format of the dataset is Chinese text, and the following content is the translation content. The format of each statement is as follows, which are stored in the . json file. The dataset is divided into training set, validation set and test set.

{
      "text": "The volcanic rocks are mainly exposed in the Tuojiqubuqu area.",
      "triple_list": [
       [
             "The volcanic rocks",
             "Exposed in",
             "The Tuojiqubuqu area"
      ] 
   ]
 }

Usage

1. Training
       python train.py
   

2. Inference
       python test.py