CCNet: Extracting High Quality Monolingual Datasets from Web Crawl Data
This work addresses the need for large, high-quality pretraining corpora in NLP, though it is incremental as it builds on existing data processing methods.
The authors tackled the problem of extracting high-quality monolingual datasets from web crawl data by developing an automatic pipeline that filters documents to resemble high-quality corpora like Wikipedia, resulting in massive datasets for multiple languages.
Pre-training text representations have led to significant improvements in many areas of natural language processing. The quality of these models benefits greatly from the size of the pretraining corpora as long as its quality is preserved. In this paper, we describe an automatic pipeline to extract massive high-quality monolingual datasets from Common Crawl for a variety of languages. Our pipeline follows the data processing introduced in fastText (Mikolov et al., 2017; Grave et al., 2018), that deduplicates documents and identifies their language. We augment this pipeline with a filtering step to select documents that are close to high quality corpora like Wikipedia.