{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:2YLB5MEYGSQB3DITDEMOJ26UFO","short_pith_number":"pith:2YLB5MEY","schema_version":"1.0","canonical_sha256":"d6161eb09834a01d8d131918e4ebd42b909503b2238218d7cb1557cc10f5b4cf","source":{"kind":"arxiv","id":"1903.04561","version":2},"attestation_state":"computed","paper":{"title":"Nuanced Metrics for Measuring Unintended Bias with Real Data for Text Classification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Daniel Borkan, Jeffrey Sorensen, Lucas Dixon, Lucy Vasserman, Nithum Thain","submitted_at":"2019-03-11T19:45:54Z","abstract_excerpt":"Unintended bias in Machine Learning can manifest as systemic differences in performance for different demographic groups, potentially compounding existing challenges to fairness in society at large. In this paper, we introduce a suite of threshold-agnostic metrics that provide a nuanced view of this unintended bias, by considering the various ways that a classifier's score distribution can vary across designated groups. We also introduce a large new test set of online comments with crowd-sourced annotations for identity references. We use this to show how our metrics can be used to find new an"},"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":"1903.04561","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-03-11T19:45:54Z","cross_cats_sorted":["cs.CL","stat.ML"],"title_canon_sha256":"9522067d8d0c5db762da0c4d1ed94eda08b9fbae99bca9c93dc226b0c8b055b7","abstract_canon_sha256":"3a612b220059874668b9d38e4e645e058c29ba797a3d7cf08b305d64e6a1fdc4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:46:45.673896Z","signature_b64":"ZCJlL44qQ94Q9RFJSTPjI+XQflZCHccsWEPZ7kLh00d213x/J/OXQDA/Pr3yuWj/5pVhKtLeri8fx0VAbDbNDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d6161eb09834a01d8d131918e4ebd42b909503b2238218d7cb1557cc10f5b4cf","last_reissued_at":"2026-05-17T23:46:45.673476Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:46:45.673476Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Nuanced Metrics for Measuring Unintended Bias with Real Data for Text Classification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Daniel Borkan, Jeffrey Sorensen, Lucas Dixon, Lucy Vasserman, Nithum Thain","submitted_at":"2019-03-11T19:45:54Z","abstract_excerpt":"Unintended bias in Machine Learning can manifest as systemic differences in performance for different demographic groups, potentially compounding existing challenges to fairness in society at large. In this paper, we introduce a suite of threshold-agnostic metrics that provide a nuanced view of this unintended bias, by considering the various ways that a classifier's score distribution can vary across designated groups. We also introduce a large new test set of online comments with crowd-sourced annotations for identity references. We use this to show how our metrics can be used to find new an"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1903.04561","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"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":"1903.04561","created_at":"2026-05-17T23:46:45.673534+00:00"},{"alias_kind":"arxiv_version","alias_value":"1903.04561v2","created_at":"2026-05-17T23:46:45.673534+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1903.04561","created_at":"2026-05-17T23:46:45.673534+00:00"},{"alias_kind":"pith_short_12","alias_value":"2YLB5MEYGSQB","created_at":"2026-05-18T12:33:07.085635+00:00"},{"alias_kind":"pith_short_16","alias_value":"2YLB5MEYGSQB3DIT","created_at":"2026-05-18T12:33:07.085635+00:00"},{"alias_kind":"pith_short_8","alias_value":"2YLB5MEY","created_at":"2026-05-18T12:33:07.085635+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.17187","citing_title":"PluRule: A Benchmark for Moderating Pluralistic Communities on Social Media","ref_index":12,"is_internal_anchor":true},{"citing_arxiv_id":"2212.04089","citing_title":"Editing Models with Task Arithmetic","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2305.10403","citing_title":"PaLM 2 Technical Report","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2YLB5MEYGSQB3DITDEMOJ26UFO","json":"https://pith.science/pith/2YLB5MEYGSQB3DITDEMOJ26UFO.json","graph_json":"https://pith.science/api/pith-number/2YLB5MEYGSQB3DITDEMOJ26UFO/graph.json","events_json":"https://pith.science/api/pith-number/2YLB5MEYGSQB3DITDEMOJ26UFO/events.json","paper":"https://pith.science/paper/2YLB5MEY"},"agent_actions":{"view_html":"https://pith.science/pith/2YLB5MEYGSQB3DITDEMOJ26UFO","download_json":"https://pith.science/pith/2YLB5MEYGSQB3DITDEMOJ26UFO.json","view_paper":"https://pith.science/paper/2YLB5MEY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1903.04561&json=true","fetch_graph":"https://pith.science/api/pith-number/2YLB5MEYGSQB3DITDEMOJ26UFO/graph.json","fetch_events":"https://pith.science/api/pith-number/2YLB5MEYGSQB3DITDEMOJ26UFO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2YLB5MEYGSQB3DITDEMOJ26UFO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2YLB5MEYGSQB3DITDEMOJ26UFO/action/storage_attestation","attest_author":"https://pith.science/pith/2YLB5MEYGSQB3DITDEMOJ26UFO/action/author_attestation","sign_citation":"https://pith.science/pith/2YLB5MEYGSQB3DITDEMOJ26UFO/action/citation_signature","submit_replication":"https://pith.science/pith/2YLB5MEYGSQB3DITDEMOJ26UFO/action/replication_record"}},"created_at":"2026-05-17T23:46:45.673534+00:00","updated_at":"2026-05-17T23:46:45.673534+00:00"}