Manually Aligned News Documents in Sinhala Tamil and English

Gold-standard benchmark for document alignment between Sinhala, Tamil, and English languages. It contains manually annotated document pairs crawled from four Sri Lankan news websites: Army, Hiru, ITN, and Newsfirst.

Use Cases

• Train a multilingual document alignment model using manually annotated Sinhala-English-Tamil pairs.
• Benchmark cross-lingual retrieval systems on gold-standard Sinhala, Tamil, and English news documents.
• Analyze linguistic patterns and translation quality across Sinhala, Tamil, and English news sources.

Strengths

• Gold-standard benchmark dataset with manual annotation for alignment quality.
• Covers three languages: Sinhala, Tamil, and English.
• Data sourced from four distinct Sri Lankan news websites.

Limitations

• Unknown dataset size, row count, and file formats.
• Geographic and topical bias towards news content from specific Sri Lankan sources.
• Potential temporal staleness as the specific crawl dates are not provided.

Provenance

Source

Crawled from Army, Hiru, ITN, and Newsfirst news websites.

Collection Method

Web crawling followed by manual annotation for document alignment.

Freshness

Last updated on 2024-02-16.

Geography

Sri Lanka (based on news source domains).