DCAD-2000: A Multilingual Dataset across 2000+ Languages with Data Cleaning as Anomaly Detection

TL;DR

This paper introduces DCAD-2000, a large-scale multilingual dataset of over 2,282 languages, constructed via anomaly detection-based data cleaning, significantly improving data quality for multilingual LLM training.

Contribution

We present a novel data cleaning approach using anomaly detection for constructing a high-quality multilingual dataset covering 2,282 languages.

Findings

Enhanced data quality and robustness demonstrated by fine-tuning LLMs on DCAD-2000.

Significant performance improvements on multilingual benchmarks, especially for low-resource languages.

Effective automatic removal of noisy data through anomaly detection-based filtering.

Abstract

The rapid development of multilingual large language models (LLMs) highlights the need for high-quality, diverse, and well-curated multilingual datasets. In this paper, we introduce DCAD-2000 (Data Cleaning as Anomaly Detection), a large-scale multilingual corpus constructed from newly extracted Common Crawl data and existing multilingual sources. DCAD-2000 covers 2,282 languages, 46.72TB of text, and 8.63 billion documents, spanning 155 high- and medium-resource languages and 159 writing scripts. To overcome the limitations of existing data cleaning approaches, which rely on manually designed heuristic thresholds, we reframe data cleaning as an anomaly detection problem. This dynamic filtering paradigm substantially improves data quality by automatically identifying and removing noisy or anomalous content. By fine-tuning LLMs on DCAD-2000, we demonstrate notable improvements in data quality, robustness of the cleaning pipeline, and downstream performance, particularly for low-resource languages across multiple multilingual benchmarks.