metadata

language:
  - ps
license:
  - cc-by-sa-4.0
size_categories:
  - 10K<n<100K

Dataset Card for Pashto Wikipedia Clean Corpus

Dataset Description

This dataset is a cleaned version of the Pashto Wikipedia dump. It has been processed to remove HTML tags, URLs, special characters, and numbers. The text has been normalized and filtered to remove very short entries.

Languages

The dataset is in Pashto (ps).

Dataset Structure

Data Instances

Each instance in the dataset contains the following fields:

url: A string of cleaned Pashto text url.
title: A string of cleaned Pashto text title.
text: A string of cleaned Pashto text.

Data Fields

url: A string of cleaned Pashto text url.
title: A string of cleaned Pashto text title.
text: A string of cleaned Pashto text.

Data Splits

The dataset contains only a train split.

Dataset Creation

Source Data

The source data is the Pashto Wikipedia dump from Hugging Face datasets.

Preprocessing

The following preprocessing steps were applied:

Removal of HTML tags
Removal of URLs
Removal of special characters and numbers
Unicode normalization
Removal of extra whitespace
Filtering out entries with less than 5 words

Considerations for Using the Data

This dataset is intended for use in language modeling and other NLP tasks for the Pashto language. Users should be aware that while efforts have been made to clean the data, some noise or errors may still be present.

Additional Information

This dataset was created and uploaded by Hanif Rahman on 21 July 2024.