{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:HE4KKV7U74J4N7Z6LK27XAA63T","short_pith_number":"pith:HE4KKV7U","schema_version":"1.0","canonical_sha256":"3938a557f4ff13c6ff3e5ab5fb801edcd99bc00ef468d57ee85d92f5488cd88f","source":{"kind":"arxiv","id":"2102.10684","version":1},"attestation_state":"computed","paper":{"title":"Pre-Training BERT on Arabic Tweets: Practical Considerations","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ahmed Abdelali, Hamdy Mubarak, Kareem Darwish, Sabit Hassan, Younes Samih","submitted_at":"2021-02-21T20:51:33Z","abstract_excerpt":"Pretraining Bidirectional Encoder Representations from Transformers (BERT) for downstream NLP tasks is a non-trival task. We pretrained 5 BERT models that differ in the size of their training sets, mixture of formal and informal Arabic, and linguistic preprocessing. All are intended to support Arabic dialects and social media. The experiments highlight the centrality of data diversity and the efficacy of linguistically aware segmentation. They also highlight that more data or more training step do not necessitate better models. Our new models achieve new state-of-the-art results on several dow"},"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":"2102.10684","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2021-02-21T20:51:33Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1d262c8465e4145acb381010fe136f4f5eec0b5bad06eca17e45267b69f920e2","abstract_canon_sha256":"da2a0167b43fe44d160d20ecbdeaa0c87c7f8654ade59a0717786282f934ea27"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:17:00.480952Z","signature_b64":"tVFPGpeqocfFF7en5wjuajIYyz/A7+BPc8Ry/5ZkvzZ564lLe0UCmSRTFB2j57/b/eFEvhAxXV5HIAHPGqnGDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3938a557f4ff13c6ff3e5ab5fb801edcd99bc00ef468d57ee85d92f5488cd88f","last_reissued_at":"2026-07-05T02:17:00.480481Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:17:00.480481Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Pre-Training BERT on Arabic Tweets: Practical Considerations","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ahmed Abdelali, Hamdy Mubarak, Kareem Darwish, Sabit Hassan, Younes Samih","submitted_at":"2021-02-21T20:51:33Z","abstract_excerpt":"Pretraining Bidirectional Encoder Representations from Transformers (BERT) for downstream NLP tasks is a non-trival task. We pretrained 5 BERT models that differ in the size of their training sets, mixture of formal and informal Arabic, and linguistic preprocessing. All are intended to support Arabic dialects and social media. The experiments highlight the centrality of data diversity and the efficacy of linguistically aware segmentation. They also highlight that more data or more training step do not necessitate better models. Our new models achieve new state-of-the-art results on several dow"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.10684","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/2102.10684/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":"2102.10684","created_at":"2026-07-05T02:17:00.480546+00:00"},{"alias_kind":"arxiv_version","alias_value":"2102.10684v1","created_at":"2026-07-05T02:17:00.480546+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.10684","created_at":"2026-07-05T02:17:00.480546+00:00"},{"alias_kind":"pith_short_12","alias_value":"HE4KKV7U74J4","created_at":"2026-07-05T02:17:00.480546+00:00"},{"alias_kind":"pith_short_16","alias_value":"HE4KKV7U74J4N7Z6","created_at":"2026-07-05T02:17:00.480546+00:00"},{"alias_kind":"pith_short_8","alias_value":"HE4KKV7U","created_at":"2026-07-05T02:17:00.480546+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25495","citing_title":"Spam and Sentiment Detection in Arabic Tweets Using MARBERT Model","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HE4KKV7U74J4N7Z6LK27XAA63T","json":"https://pith.science/pith/HE4KKV7U74J4N7Z6LK27XAA63T.json","graph_json":"https://pith.science/api/pith-number/HE4KKV7U74J4N7Z6LK27XAA63T/graph.json","events_json":"https://pith.science/api/pith-number/HE4KKV7U74J4N7Z6LK27XAA63T/events.json","paper":"https://pith.science/paper/HE4KKV7U"},"agent_actions":{"view_html":"https://pith.science/pith/HE4KKV7U74J4N7Z6LK27XAA63T","download_json":"https://pith.science/pith/HE4KKV7U74J4N7Z6LK27XAA63T.json","view_paper":"https://pith.science/paper/HE4KKV7U","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2102.10684&json=true","fetch_graph":"https://pith.science/api/pith-number/HE4KKV7U74J4N7Z6LK27XAA63T/graph.json","fetch_events":"https://pith.science/api/pith-number/HE4KKV7U74J4N7Z6LK27XAA63T/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HE4KKV7U74J4N7Z6LK27XAA63T/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HE4KKV7U74J4N7Z6LK27XAA63T/action/storage_attestation","attest_author":"https://pith.science/pith/HE4KKV7U74J4N7Z6LK27XAA63T/action/author_attestation","sign_citation":"https://pith.science/pith/HE4KKV7U74J4N7Z6LK27XAA63T/action/citation_signature","submit_replication":"https://pith.science/pith/HE4KKV7U74J4N7Z6LK27XAA63T/action/replication_record"}},"created_at":"2026-07-05T02:17:00.480546+00:00","updated_at":"2026-07-05T02:17:00.480546+00:00"}