Welcome to KAUST and Iraya's virtual summer school on Utilising unstructured data in geoscience. Over three days, we will give you an introduction to what is Natural Language Processing, how the geoscience industry is leveraging it, and the new avenues that are up for exploration.
In this repository we have shared our codes for the tutorials as well as guidance on setting this up on your workstation. The lectures which accompany the tutorials will be available as video recordings in the near future.
To create the environment which has all the necessary python packages pre-installed, in your terminal in this folder directory run:
conda env create -f environment_dayX.yml
This creates a conda environment named "mlg_summer_school". Before we begin you will need to activate this environment. To do so, in your terminal run the command:
conda activate mlg_summer_school_dayX
Where in both cases X=1,2,3. Note that for day3 you will need a machine with a GPU. Otherwise simply use Colab.
All participants are expected to have a basic knowledge of python.
| Day | Tutorial (Github) | Tutorial (Colab) | Videos |
|---|---|---|---|
| 1: Introduction to NLP | Link |  |
Link |
| 2: Active Learning | Link | |
Link |
| 3a: BERT | Link | |
Link |
| 3b: Title Generation | Link | |
- |
The sample test data used for this experiment is extracted from ED2K, an initiative between Iraya Energies and European Association of Geoscientists and Engineer (EAGE) to process the EarthDoc database, using the latest in machine learning techniques to read the earth better.
The API access will be available until 17th June 2021 and is available for academic use only. Thank you in advance for deletion of downloaded dataset after the completion of the class. For the utilization of data beyond the duration of this Summer School or use of bigger ED2K dataset, pls send an email request to info@irayaenergies.com.
This virtual school has been organised in collaboration between King Abdullah University of Science and Technology (KAUST) and Iraya Energies.