{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:H2PBIZ5RZLSLH7C6RTGZKYZC4D","short_pith_number":"pith:H2PBIZ5R","schema_version":"1.0","canonical_sha256":"3e9e1467b1cae4b3fc5e8ccd956322e0d1a0a10be97ca64b0643150800f31451","source":{"kind":"arxiv","id":"2203.10258","version":3},"attestation_state":"computed","paper":{"title":"TDR-CL: Targeted Doubly Robust Collaborative Learning for Debiased Recommendations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.IR","authors_text":"Chunyuan Zheng, Haoxuan Li, Peng Wu, Yan Lyu","submitted_at":"2022-03-19T06:48:50Z","abstract_excerpt":"Bias is a common problem inherent in recommender systems, which is entangled with users' preferences and poses a great challenge to unbiased learning. For debiasing tasks, the doubly robust (DR) method and its variants show superior performance due to the double robustness property, that is, DR is unbiased when either imputed errors or learned propensities are accurate. However, our theoretical analysis reveals that DR usually has a large variance. Meanwhile, DR would suffer unexpectedly large bias and poor generalization caused by inaccurate imputed errors and learned propensities, which usua"},"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":"2203.10258","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2022-03-19T06:48:50Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"74c23071d6163870575a2d8c4462860a42be915ee0cc0aa162ac44f8584e5e4b","abstract_canon_sha256":"1197e8a78ada7b86db37d04a77879eada62ab4e1d3df6b15af9ac542e1305b24"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:47:14.952239Z","signature_b64":"Ky1peQkmO3zn1VEbXvXP8LUYtvIkQ0TLbhNTIcRjEVD4cl5t0ccLNauOXPvPKE3xrZpN7oc+CEszGFltx+5dAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3e9e1467b1cae4b3fc5e8ccd956322e0d1a0a10be97ca64b0643150800f31451","last_reissued_at":"2026-07-05T05:47:14.951895Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:47:14.951895Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TDR-CL: Targeted Doubly Robust Collaborative Learning for Debiased Recommendations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.IR","authors_text":"Chunyuan Zheng, Haoxuan Li, Peng Wu, Yan Lyu","submitted_at":"2022-03-19T06:48:50Z","abstract_excerpt":"Bias is a common problem inherent in recommender systems, which is entangled with users' preferences and poses a great challenge to unbiased learning. For debiasing tasks, the doubly robust (DR) method and its variants show superior performance due to the double robustness property, that is, DR is unbiased when either imputed errors or learned propensities are accurate. However, our theoretical analysis reveals that DR usually has a large variance. Meanwhile, DR would suffer unexpectedly large bias and poor generalization caused by inaccurate imputed errors and learned propensities, which usua"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.10258","kind":"arxiv","version":3},"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/2203.10258/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":"2203.10258","created_at":"2026-07-05T05:47:14.951951+00:00"},{"alias_kind":"arxiv_version","alias_value":"2203.10258v3","created_at":"2026-07-05T05:47:14.951951+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.10258","created_at":"2026-07-05T05:47:14.951951+00:00"},{"alias_kind":"pith_short_12","alias_value":"H2PBIZ5RZLSL","created_at":"2026-07-05T05:47:14.951951+00:00"},{"alias_kind":"pith_short_16","alias_value":"H2PBIZ5RZLSLH7C6","created_at":"2026-07-05T05:47:14.951951+00:00"},{"alias_kind":"pith_short_8","alias_value":"H2PBIZ5R","created_at":"2026-07-05T05:47:14.951951+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.24092","citing_title":"ConAlign: Conditional Alignment Framework for Balancing Biased and Unbiased Recommendation","ref_index":13,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/H2PBIZ5RZLSLH7C6RTGZKYZC4D","json":"https://pith.science/pith/H2PBIZ5RZLSLH7C6RTGZKYZC4D.json","graph_json":"https://pith.science/api/pith-number/H2PBIZ5RZLSLH7C6RTGZKYZC4D/graph.json","events_json":"https://pith.science/api/pith-number/H2PBIZ5RZLSLH7C6RTGZKYZC4D/events.json","paper":"https://pith.science/paper/H2PBIZ5R"},"agent_actions":{"view_html":"https://pith.science/pith/H2PBIZ5RZLSLH7C6RTGZKYZC4D","download_json":"https://pith.science/pith/H2PBIZ5RZLSLH7C6RTGZKYZC4D.json","view_paper":"https://pith.science/paper/H2PBIZ5R","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2203.10258&json=true","fetch_graph":"https://pith.science/api/pith-number/H2PBIZ5RZLSLH7C6RTGZKYZC4D/graph.json","fetch_events":"https://pith.science/api/pith-number/H2PBIZ5RZLSLH7C6RTGZKYZC4D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/H2PBIZ5RZLSLH7C6RTGZKYZC4D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/H2PBIZ5RZLSLH7C6RTGZKYZC4D/action/storage_attestation","attest_author":"https://pith.science/pith/H2PBIZ5RZLSLH7C6RTGZKYZC4D/action/author_attestation","sign_citation":"https://pith.science/pith/H2PBIZ5RZLSLH7C6RTGZKYZC4D/action/citation_signature","submit_replication":"https://pith.science/pith/H2PBIZ5RZLSLH7C6RTGZKYZC4D/action/replication_record"}},"created_at":"2026-07-05T05:47:14.951951+00:00","updated_at":"2026-07-05T05:47:14.951951+00:00"}