Cleaned_Dataset.csv – The combined CSV files of all scraped documents from DABI, e-LiS, o-bib and Springer.

Data_Cleaning.ipynb – The Jupyter Notebook with python code for the analysis and cleaning of the original dataset.

ger_train.csv – The German training set as CSV file.

ger_validation.csv – The German validation set as CSV file.

en_test.csv – The English test set as CSV file.

en_train.csv – The English training set as CSV file.

en_validation.csv – The English validation set as CSV file.

splitting.py – The python code for splitting a dataset into train, test and validation set.

DataSetTrans_de.csv – The final German dataset as a CSV file.

DataSetTrans_en.csv – The final English dataset as a CSV file.

translation.py – The python code for translating the cleaned dataset.

This repository accompanies the manuscript "Spatially resolved uncertainties for machine learning potentials" by E. Heid, J. Schörghuber, R. Wanzenböck, and G. K. H. Madsen. The following files are available:

• mc_experiment.ipynb is a Jupyter notebook for the Monte Carlo experiment described in the study (artificial model with only variance as error source).

• aggregate_cut_relax.py contains code to cut and relax boxes for the water active learning cycle.

• data_t1x.tar.gz contains reaction pathways for 10,073 reactions from a subset of the Transition1x dataset, split into training, validation and test sets. The training and validation sets contain the indices 1, 2, 9, and 10 from a 10-image nudged-elastic band search (40k datapoints), while the test set contains indices 3-8 (60k datapoints). The test set is ordered according to the reaction and index, i.e. rxn1_index3, rxn1_index4, [...] rxn1_index8, rxn2_index3, [...].

• data_sto.tar.gz contains surface reconstructions of SrTiO3, randomly split into a training and validation set, as well as a test set.

• data_h2o.tar.gz contains:

  • full_db.extxyz: The full dataset of 1.5k structures.

  • iter00_train.extxyz and iter00_validation.extxyz: The initial training and validation set for the active learning cycle.

  • the subfolders in the folders random and uncertain contain the training and validation sets for the random and uncertainty-based active learning loops.

The Blog-1K corpus is a redistributable authorship identification testbed for contemporary English prose. It has 1,000 candidate authors, 16K+ posts, and a pre-defined data split (train/dev/test proportional to ca. 8:1:1). It is a subset of the Blog Authorship Corpus from Kaggle. The MD5 for Blog-1K is '0a9e38740af9f921b6316b7f400acf06'.

1. Preprocessing

We first filter out texts shorter than 1,000 characters. Then we select one thousand authors whose writings meet the following criteria: - accumulatively at least 10,000 characters, - accumulatively at most 49,410 characters, - accumulatively at least 16 posts, - accumulatively at most 40 posts, and - each text has at least 50 function words found in the Koppel512 list (to filter out non-English prose).

Blog-1K has three columns: 'id', 'text', and 'split', where 'id' corresponds to its parent corpus.

1. Statistics

Its creation and statistics can be found in the Jupyter Notebook.

    Split
    # Authors
    # Posts
    # Characters
    Avg. Characters Per Author (Std.)
    Avg. Characters Per Post (Std.)


    Train
    1,000
    16,132
    30,092,057
    30,092 (5,884)
    1,865 (1,007)


    Validation
    935
    2,017
    3,755,362
    4,016 (2,269)
    1,862 (999)


    Test
    924
    2,017
    3,732,448
    4,039 (2,188)
    1,850 (936)

1. Usage

import pandas as pd

df = pd.read_csv('blog1000.csv.gz', compression='infer')

read in training data

train_text, train_label = zip(*df.loc[df.split=='train'][['text', 'id']].itertuples(index=False))

1. License All the materials is licensed under the ISC License.

2. Contact Please contact its maintainer for questions.

Release of the experimental data from the paper Towards Linking Graph Topology to Model Performance for Biomedical Knowledge Graph Completion (accepted at Machine Learning for Life and Material Sciences workshop @ ICML2024).