{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:WQSZ647UJNNCVR43RL6I2AXCZD","short_pith_number":"pith:WQSZ647U","schema_version":"1.0","canonical_sha256":"b4259f73f44b5a2ac79b8afc8d02e2c8ec37c6a33cf52720b3cd53d50b6e884b","source":{"kind":"arxiv","id":"2507.20227","version":3},"attestation_state":"computed","paper":{"title":"CTR-Driven Ad Text Generation via Online Feedback Preference Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Bo Zheng, Jiale Chen, Qixiang Gao, Si Chen, Tiezheng Ge, Xubin Li, Yanda Chen, Zihui Ren","submitted_at":"2025-07-27T11:13:03Z","abstract_excerpt":"Advertising text plays a critical role in determining click-through rates (CTR) in online advertising. Large Language Models (LLMs) offer significant efficiency advantages over manual ad text creation. However, LLM-generated ad texts do not guarantee higher CTR performance compared to human-crafted texts, revealing a gap between generation quality and online performance of ad texts. In this work, we propose a novel ad text generation method which optimizes for CTR through preference optimization from online feedback. Our approach adopts an innovative two-stage framework: (1) diverse ad text sa"},"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":"2507.20227","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2025-07-27T11:13:03Z","cross_cats_sorted":[],"title_canon_sha256":"711410c3b71da84fc6dab1144832ca6e8a6984db50a7fca6c792766ee30ce37d","abstract_canon_sha256":"e60710c91ec09c7e4a3d6ba9a523ec550192cbb921670056563fcccebeeffe6a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:47:27.007966Z","signature_b64":"l0rsV/KTLIEQA7GfhrzFwCEHdCh95vat+IQEqTcdOXLKH+aPQ1E5tppvdJ4wSb/GuEUnHwIGIln+MJGyCuPaDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b4259f73f44b5a2ac79b8afc8d02e2c8ec37c6a33cf52720b3cd53d50b6e884b","last_reissued_at":"2026-07-05T11:47:27.007468Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:47:27.007468Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CTR-Driven Ad Text Generation via Online Feedback Preference Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Bo Zheng, Jiale Chen, Qixiang Gao, Si Chen, Tiezheng Ge, Xubin Li, Yanda Chen, Zihui Ren","submitted_at":"2025-07-27T11:13:03Z","abstract_excerpt":"Advertising text plays a critical role in determining click-through rates (CTR) in online advertising. Large Language Models (LLMs) offer significant efficiency advantages over manual ad text creation. However, LLM-generated ad texts do not guarantee higher CTR performance compared to human-crafted texts, revealing a gap between generation quality and online performance of ad texts. In this work, we propose a novel ad text generation method which optimizes for CTR through preference optimization from online feedback. Our approach adopts an innovative two-stage framework: (1) diverse ad text sa"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.20227","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/2507.20227/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":"2507.20227","created_at":"2026-07-05T11:47:27.007530+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.20227v3","created_at":"2026-07-05T11:47:27.007530+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.20227","created_at":"2026-07-05T11:47:27.007530+00:00"},{"alias_kind":"pith_short_12","alias_value":"WQSZ647UJNNC","created_at":"2026-07-05T11:47:27.007530+00:00"},{"alias_kind":"pith_short_16","alias_value":"WQSZ647UJNNCVR43","created_at":"2026-07-05T11:47:27.007530+00:00"},{"alias_kind":"pith_short_8","alias_value":"WQSZ647U","created_at":"2026-07-05T11:47:27.007530+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.12138","citing_title":"Design Your Ad: Personalized Advertising Image and Text Generation with Unified Autoregressive Models","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WQSZ647UJNNCVR43RL6I2AXCZD","json":"https://pith.science/pith/WQSZ647UJNNCVR43RL6I2AXCZD.json","graph_json":"https://pith.science/api/pith-number/WQSZ647UJNNCVR43RL6I2AXCZD/graph.json","events_json":"https://pith.science/api/pith-number/WQSZ647UJNNCVR43RL6I2AXCZD/events.json","paper":"https://pith.science/paper/WQSZ647U"},"agent_actions":{"view_html":"https://pith.science/pith/WQSZ647UJNNCVR43RL6I2AXCZD","download_json":"https://pith.science/pith/WQSZ647UJNNCVR43RL6I2AXCZD.json","view_paper":"https://pith.science/paper/WQSZ647U","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.20227&json=true","fetch_graph":"https://pith.science/api/pith-number/WQSZ647UJNNCVR43RL6I2AXCZD/graph.json","fetch_events":"https://pith.science/api/pith-number/WQSZ647UJNNCVR43RL6I2AXCZD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WQSZ647UJNNCVR43RL6I2AXCZD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WQSZ647UJNNCVR43RL6I2AXCZD/action/storage_attestation","attest_author":"https://pith.science/pith/WQSZ647UJNNCVR43RL6I2AXCZD/action/author_attestation","sign_citation":"https://pith.science/pith/WQSZ647UJNNCVR43RL6I2AXCZD/action/citation_signature","submit_replication":"https://pith.science/pith/WQSZ647UJNNCVR43RL6I2AXCZD/action/replication_record"}},"created_at":"2026-07-05T11:47:27.007530+00:00","updated_at":"2026-07-05T11:47:27.007530+00:00"}