MegaWika 2: A More Comprehensive Multilingual Collection of Articles and their Sources

2025/08/05 by Barham, Samuel, May, Chandler, Van Durme, Benjamin · 1 citation
#Computation and Language (cs.CL) #Digital Libraries (cs.DL) #FOS: Computer and information sciences

paper · doi:10.48550/arxiv.2508.03828

Abstract

We introduce MegaWika 2, a large, multilingual dataset of Wikipedia articles with their citations and scraped web sources; articles are represented in a rich data structure, and scraped source texts are stored inline with precise character offsets of their citations in the article text. MegaWika 2 is a major upgrade from the original MegaWika, spanning six times as many articles and twice as many fully scraped citations. Both MegaWika and MegaWika 2 support report generation research ; whereas MegaWika also focused on supporting question answering and retrieval applications, MegaWika 2 is designed to support fact checking and analyses across time and language.

Citations

• Entity Insertion in Multilingual Linked Corpora: The Case of Wikipedia
• GrowOVER: How Can LLMs Adapt to Growing Real-World Knowledge?
• AdapterSwap: Continuous Training of LLMs with Data Removal and Access-Control Guarantees
• Simpson's Paradox and the Accuracy-Fluency Tradeoff in Translation
• Multilingual E5 Text Embeddings: A Technical Report
• OLMo: Accelerating the Science of Language Models
• The Falcon Series of Open Language Models
• Carpe Diem: On the Evaluation of World Knowledge in Lifelong Language Models
• TempTabQA: Temporal Question Answering for Semi-Structured Tables
• Leveraging Recommender Systems to Reduce Content Gaps on Peer Production Platforms
• MegaWika: Millions of reports and their sources across 50 diverse languages
• Open-WikiTable: Dataset for Open Domain Question Answering with Complex Reasoning over Table
• Evaluating Verifiability in Generative Search Engines
• GPT-4 Technical Report
• WiCE: Real-World Entailment for Claims in Wikipedia
• LLaMA: Open and Efficient Foundation Language Models
• Making a MIRACL: Multilingual Information Retrieval Across a Continuum of Languages
• No Language Left Behind: Scaling Human-Centered Machine Translation
• OPT: Open Pre-trained Transformer Language Models
• PaLM: Scaling Language Modeling with Pathways
• The Web Is Your Oyster - Knowledge-Intensive NLP against a Very Large Web Corpus
• Scaling Language Models: Methods, Analysis & Insights from Training Gopher
• WikiCheck: An end-to-end open source Automatic Fact-Checking API based\n on Wikipedia
• Mr. TyDi: A Multi-lingual Benchmark for Dense Retrieval
• A Multilingual Entity Linking System for Wikipedia with a Machine-in-the-Loop Approach
• PAQ: 65 Million Probably-Asked Questions and What You Can Do With Them
• WikiTableT: A Large-Scale Data-to-Text Dataset for Generating Wikipedia Article Sections
• XOR QA: Cross-lingual Open-Retrieval Question Answering
• Beyond English-Centric Multilingual Machine Translation
• MKQA: A Linguistically Diverse Benchmark for Multilingual Open Domain Question Answering
• Language Models are Few-Shot Learners
• ToTTo: A Controlled Table-To-Text Generation Dataset
• Logical Natural Language Generation from Open-Domain Tables
• TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages
• Unsupervised Cross-lingual Representation Learning at Scale
• On the Cross-lingual Transferability of Monolingual Representations
• MLQA: Evaluating Cross-lingual Extractive Question Answering
• TabFact: A Large-scale Dataset for Table-based Fact Verification
• RoBERTa: A Robustly Optimized BERT Pretraining Approach
• Generating Summaries with Topic Templates and Structured Convolutional Decoders
• Why the World Reads Wikipedia: Beyond English Speakers
• Harvesting Paragraph-Level Question-Answer Pairs from Wikipedia
• Generating Wikipedia by Summarizing Long Sequences
• Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learning
• Reading Wikipedia to Answer Open-Domain Questions
• Pointer Sentinel Mixture Models

Cited by

• mmBERT: A Modern Multilingual Encoder with Annealed Language Learning

Related

• MegaWika: Millions of reports and their sources across 50 diverse languages
• Wikipedia Citations: Reproducible Citation Extraction from Multilingual Wikipedia
• An Open Multilingual System for Scoring Readability of Wikipedia
• Wikipedia Citations: A comprehensive dataset of citations with identifiers extracted from English Wikipedia
• WikiHist.html: English Wikipedia's Full Revision History in HTML Format