Philippine English Text Corpus from News, Religion, and Legal Domains

PhEnText is a lexical dataset of Philippine English text containing 20,562,265 lines sourced from news articles, religious articles, and court decisions. It was created by joypersicanon and last updated in March 2022.

Use Cases

• Analyze lexical patterns and domain-specific vocabulary across news, religious, and legal text lines.
• Train language models on Philippine English text using the 20.5 million lines for tasks like text generation.
• Study the distribution and frequency of terms within the three specified domains (news, religion, legal) present in the corpus.

Strengths

• Large-scale corpus with 20,562,265 lines of text.
• Multi-domain coverage includes news, religious, and legal documents.

Limitations

• No column definitions or sample data are available to understand the data structure.
• The dataset's recency is limited, with a last update in March 2022.

Provenance

Source

huggingface

Collection Method

Aggregated from news articles, religious articles, and court decisions.

Freshness

Last updated 2022-03-17.

Geography

Philippines (implied by 'Philippine English')