cleaning_metadata.json

{
  "dataset_info": {
    "source": "Enron Email Dataset (cleaned)",
    "processed_date": "2025-06-22T20:30:02.258451",
    "cleaning_applied": true,
    "original_files": [
      "C:\\Users\\jvina\\Documents\\Curso 2024-25\\TFG\\Repositorio\\Unlearning\\pretraining\\datasets\\emails_phillip_allen.jsonl",
      "C:\\Users\\jvina\\Documents\\Curso 2024-25\\TFG\\Repositorio\\Unlearning\\pretraining\\datasets\\emails_others_part1.jsonl"
    ],
    "columns_kept": [
      "text"
    ],
    "format": "JSONL with single 'text' column"
  },
  "phillip_allen": {
    "original_count": 1423,
    "cleaned_count": 319,
    "removed_count": 1104,
    "removal_percentage": 77.58257203092059,
    "output_file": "C:\\Users\\jvina\\Documents\\Curso 2024-25\\TFG\\Repositorio\\cleaned_datasets\\emails_phillip_allen_cleaned.jsonl",
    "columns": [
      "text"
    ]
  },
  "others": {
    "original_count": 3748,
    "cleaned_count": 1663,
    "removed_count": 2085,
    "removal_percentage": 55.62966915688368,
    "output_file": "C:\\Users\\jvina\\Documents\\Curso 2024-25\\TFG\\Repositorio\\cleaned_datasets\\emails_others_cleaned.jsonl",
    "columns": [
      "text"
    ]
  },
  "cleaning_steps": [
    "Eliminación de valores NaN en columna 'text'",
    "Eliminación de textos vacíos",
    "Eliminación de strings 'nan'/'null'",
    "Filtrado por longitud (10-2000 caracteres)",
    "Limpieza de caracteres problemáticos",
    "Eliminación de duplicados exactos",
    "Validación de estructura de email",
    "Mantener solo columna 'text'"
  ]
}