Vietnamese WikiQA: 7,592 Question-Answer Pairs from Wikipedia

7,592 Vietnamese question-answer pairs were generated from Vietnamese Wikipedia sections. The dataset was created by HienNGuit for QA research, dataset analysis, and difficulty modeling, with a last recorded update in June 2026. It underwent automatic filtering and human verification before release.

Use Cases

• Training question-answering models based on Vietnamese Wikipedia content.
• Analyzing dataset difficulty for Vietnamese NLP tasks.
• Benchmarking automatic filtering and human verification processes for QA data.

Strengths

• Contains 7,592 QA pairs, providing a substantial corpus for Vietnamese NLP.
• Data underwent a multi-stage pipeline including generation, automatic filtering, and human verification.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is known, but other structural details like file formats are unspecified.

Provenance

Source

Vietnamese Wikipedia.

Collection Method

Generated from Wikipedia sections, then automatically filtered and human-verified.

Freshness

Last updated 2026-06-02 10:56:05; freshness should be verified.