MegaWika: Millions of reports and their sources across 50 diverse languages

TL;DR

MegaWika is a comprehensive multilingual dataset with 13 million Wikipedia articles and 71 million sources, enabling advanced AI-assisted report generation, translation, and semantic analysis across 50 languages.

Contribution

We introduce MegaWika, the largest multilingual report generation dataset, and provide baseline models for cross-lingual question answering and citation retrieval.

Findings

MegaWika contains 13 million articles and 71 million sources.

Baseline models demonstrate effective cross-lingual question answering.

Semantic analysis confirms dataset quality and diversity.

Abstract

To foster the development of new models for collaborative AI-assisted report generation, we introduce MegaWika, consisting of 13 million Wikipedia articles in 50 diverse languages, along with their 71 million referenced source materials. We process this dataset for a myriad of applications, going beyond the initial Wikipedia citation extraction and web scraping of content, including translating non-English articles for cross-lingual applications and providing FrameNet parses for automated semantic analysis. MegaWika is the largest resource for sentence-level report generation and the only report generation dataset that is multilingual. We manually analyze the quality of this resource through a semantically stratified sample. Finally, we provide baseline results and trained models for crucial steps in automated report generation: cross-lingual question answering and citation retrieval.