{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:6MVUZ5BLTEFTAEMXFVL72KMS7R","short_pith_number":"pith:6MVUZ5BL","schema_version":"1.0","canonical_sha256":"f32b4cf42b990b3011972d57fd2992fc47483de3c248214bcff15291b0d648f1","source":{"kind":"arxiv","id":"2201.06642","version":1},"attestation_state":"computed","paper":{"title":"Towards a Cleaner Document-Oriented Multilingual Crawled Corpus","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Beno\\^it Sagot, Julien Abadji, Laurent Romary, Pedro Ortiz Suarez","submitted_at":"2022-01-17T22:12:59Z","abstract_excerpt":"The need for raw large raw corpora has dramatically increased in recent years with the introduction of transfer learning and semi-supervised learning methods to Natural Language Processing. And while there have been some recent attempts to manually curate the amount of data necessary to train large language models, the main way to obtain this data is still through automatic web crawling. In this paper we take the existing multilingual web corpus OSCAR and its pipeline Ungoliant that extracts and classifies data from Common Crawl at the line level, and propose a set of improvements and automati"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2201.06642","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-01-17T22:12:59Z","cross_cats_sorted":[],"title_canon_sha256":"ef7cc4b7f39c669a251a732b552366ac98a507e23d2438645cb5046d61ceb9c1","abstract_canon_sha256":"ecb2aae68c4c128d6f2cf0058bbeb286ee1c6492a3935947385fae273073bf64"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:49:00.242977Z","signature_b64":"yn9zpV6howQ9ejJfQI+8PXNbsnsZcKYHHDLC/GqDFDJxXU47JBYc1Evq2N2XuTB1Em7kV+BgnAZA+2fyI/zoBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f32b4cf42b990b3011972d57fd2992fc47483de3c248214bcff15291b0d648f1","last_reissued_at":"2026-07-05T03:49:00.242562Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:49:00.242562Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards a Cleaner Document-Oriented Multilingual Crawled Corpus","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Beno\\^it Sagot, Julien Abadji, Laurent Romary, Pedro Ortiz Suarez","submitted_at":"2022-01-17T22:12:59Z","abstract_excerpt":"The need for raw large raw corpora has dramatically increased in recent years with the introduction of transfer learning and semi-supervised learning methods to Natural Language Processing. And while there have been some recent attempts to manually curate the amount of data necessary to train large language models, the main way to obtain this data is still through automatic web crawling. In this paper we take the existing multilingual web corpus OSCAR and its pipeline Ungoliant that extracts and classifies data from Common Crawl at the line level, and propose a set of improvements and automati"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2201.06642","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2201.06642/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2201.06642","created_at":"2026-07-05T03:49:00.242619+00:00"},{"alias_kind":"arxiv_version","alias_value":"2201.06642v1","created_at":"2026-07-05T03:49:00.242619+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2201.06642","created_at":"2026-07-05T03:49:00.242619+00:00"},{"alias_kind":"pith_short_12","alias_value":"6MVUZ5BLTEFT","created_at":"2026-07-05T03:49:00.242619+00:00"},{"alias_kind":"pith_short_16","alias_value":"6MVUZ5BLTEFTAEMX","created_at":"2026-07-05T03:49:00.242619+00:00"},{"alias_kind":"pith_short_8","alias_value":"6MVUZ5BL","created_at":"2026-07-05T03:49:00.242619+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.24718","citing_title":"The Tokenizer Tax Across 25 European Languages: Domain Invariance, Cross-Lingual Few-Shot Effects, and the Ukrainian Penalty","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2311.16867","citing_title":"The Falcon Series of Open Language Models","ref_index":235,"is_internal_anchor":false},{"citing_arxiv_id":"2306.01116","citing_title":"The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora with Web Data, and Web Data Only","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2406.17557","citing_title":"The FineWeb Datasets: Decanting the Web for the Finest Text Data at Scale","ref_index":25,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6MVUZ5BLTEFTAEMXFVL72KMS7R","json":"https://pith.science/pith/6MVUZ5BLTEFTAEMXFVL72KMS7R.json","graph_json":"https://pith.science/api/pith-number/6MVUZ5BLTEFTAEMXFVL72KMS7R/graph.json","events_json":"https://pith.science/api/pith-number/6MVUZ5BLTEFTAEMXFVL72KMS7R/events.json","paper":"https://pith.science/paper/6MVUZ5BL"},"agent_actions":{"view_html":"https://pith.science/pith/6MVUZ5BLTEFTAEMXFVL72KMS7R","download_json":"https://pith.science/pith/6MVUZ5BLTEFTAEMXFVL72KMS7R.json","view_paper":"https://pith.science/paper/6MVUZ5BL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2201.06642&json=true","fetch_graph":"https://pith.science/api/pith-number/6MVUZ5BLTEFTAEMXFVL72KMS7R/graph.json","fetch_events":"https://pith.science/api/pith-number/6MVUZ5BLTEFTAEMXFVL72KMS7R/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6MVUZ5BLTEFTAEMXFVL72KMS7R/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6MVUZ5BLTEFTAEMXFVL72KMS7R/action/storage_attestation","attest_author":"https://pith.science/pith/6MVUZ5BLTEFTAEMXFVL72KMS7R/action/author_attestation","sign_citation":"https://pith.science/pith/6MVUZ5BLTEFTAEMXFVL72KMS7R/action/citation_signature","submit_replication":"https://pith.science/pith/6MVUZ5BLTEFTAEMXFVL72KMS7R/action/replication_record"}},"created_at":"2026-07-05T03:49:00.242619+00:00","updated_at":"2026-07-05T03:49:00.242619+00:00"}