Datasets:

Modalities:
Text
Formats:
parquet
DOI:
Libraries:
Datasets
pandas
License:

You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

In-depth Analysis of the Urdu Corpus Dataset

This repository contains an in-depth analysis of an Urdu corpus dataset. The goal of this analysis is to provide insights into the Urdu text, including word frequency, character distribution, and more.

Dataset Overview

  • Number of Records: 6,315
  • Total Words: 6,453,695
  • Average Words per Record: 1,021.96
  • Number of Unique Words: 187,256
  • Vocabulary Richness: 0.0290 (Unique Words / Total Words)
  • Number of Hapax Legomena: 108,026 (Words that appear only once)
  • Total Characters: 29,807,965
  • Average Characters per Record: 4,720.18
  • Total Sentences (Estimated): 230,284
  • Average Sentence Length: 28.02 words

Most Common Words

The following are the 50 most common words in the corpus:

  • کے: 247,267
  • کی: 186,780
  • میں: 175,867
  • اور: 133,344
  • سے: 124,138
  • کا: 109,264
  • اس: 95,935
  • ہے: 91,902
  • کو: 88,741
  • کہ: 85,883
  • نے: 72,281
  • پر: 62,794
  • یہ: 55,129
  • ہے۔: 53,279
  • بھی: 51,484
  • کر: 51,201
  • ایک: 44,305
  • ان: 43,296
  • نہیں: 40,128
  • تو: 36,780

... and more.

Most Common Characters

The following are the 20 most common characters in the corpus:

  • Space: 6,447,380
  • ا: 2,918,252
  • ی: 2,328,438
  • ک: 1,482,038
  • ر: 1,445,200
  • و: 1,384,407
  • ہ: 1,224,706
  • ن: 1,168,827
  • ے: 1,138,267
  • م: 1,108,999

... and more.

Record Length Analysis

  • Shortest Record (Length: 13 words):
    • "دوائیوں کا کاروبار یاسر پیرزادہ آئی بی سی اردو آئی بی سی اردو"

image/png

  • Longest Record: The longest record contains 30200 words.

image/png

Usage

You can use this dataset and analysis results for further natural language processing tasks, such as:

  • Language modeling
  • Sentiment analysis
  • Text generation
  • Translation

longest sentence sample image

Contributing

Feel free to fork this repository and contribute to improving the dataset or extending the analysis.

License

see above model card for License.

Downloads last month
30