Software Open Access

Replication package for identify bot comments

Mehdi Golzadeh; Alexandre Decan; Eleni Constantinou; Tom Mens


Citation Style Language JSON Export

{
  "publisher": "Zenodo", 
  "DOI": "10.5281/zenodo.4580998", 
  "language": "eng", 
  "title": "Replication package for identify bot comments", 
  "issued": {
    "date-parts": [
      [
        2021, 
        5, 
        22
      ]
    ]
  }, 
  "abstract": "<p>This repository contains the replication package for our study about identifying bots at the level of their activity in GitHub submitted to BotSE&#39;21 conference (*&quot;Identifying bot activity in GitHub pull request and issue comments&quot;*).<br>\nA link to the paper will be added to this README as soon as the paper is accepted.</p>\n\n<p><strong>Ground-truth dataset</strong><br>\nThe dataset is extracted from the ground-truth dataset of our study about [identifying bots](https://arxiv.org/abs/2010.03303) published in JSS journal.</p>\n\n<p><strong>Replication package</strong></p>\n\n<p>A- Dataset preparation.ipynb: This notebook splits the dataset to two disjoint set for training and test purposes. To avoid any conflict with GDPR regulations we&#39;ve anonymised the account name columns.</p>\n\n<p>B- Model construction.ipynb: We followed a Grid-search cross validation in this notebook to find the best classifier and construct the final mode. The replication package was originally created on Python 3.8&nbsp; and the dependencies required to run these notebooks are listed in requirements.txt and can be automatically installed using pip install -r requirements.txt.</p>\n\n<p>C- Model evaluation.ipynb: this notebook contains scripts to evaluate the classifier.</p>", 
  "author": [
    {
      "family": "Mehdi Golzadeh"
    }, 
    {
      "family": "Alexandre Decan"
    }, 
    {
      "family": "Eleni Constantinou"
    }, 
    {
      "family": "Tom Mens"
    }
  ], 
  "version": "1.0.0", 
  "type": "article", 
  "id": "4580998"
}
47
10
views
downloads
All versions This version
Views 4742
Downloads 1010
Data volume 92.6 MB92.6 MB
Unique views 4340
Unique downloads 1010

Share

Cite as