{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:WFROJZIGTYWHNDWFWKJFLQVJSW","short_pith_number":"pith:WFROJZIG","schema_version":"1.0","canonical_sha256":"b162e4e5069e2c768ec5b29255c2a9958937636d1071d8cf78b4ff8fc4a2b956","source":{"kind":"arxiv","id":"2505.17410","version":1},"attestation_state":"computed","paper":{"title":"LLM-based Generative Error Correction for Rare Words with Synthetic Data and Phonetic Context","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","eess.AS"],"primary_cat":"cs.SD","authors_text":"Hiroaki Kokubo, Masaaki Yamamoto, Natsuo Yamashita, Yohei Kawaguchi","submitted_at":"2025-05-23T02:54:52Z","abstract_excerpt":"Generative error correction (GER) with large language models (LLMs) has emerged as an effective post-processing approach to improve automatic speech recognition (ASR) performance. However, it often struggles with rare or domain-specific words due to limited training data. Furthermore, existing LLM-based GER approaches primarily rely on textual information, neglecting phonetic cues, which leads to over-correction. To address these issues, we propose a novel LLM-based GER approach that targets rare words and incorporates phonetic information. First, we generate synthetic data to contain rare wor"},"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":"2505.17410","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SD","submitted_at":"2025-05-23T02:54:52Z","cross_cats_sorted":["cs.CL","eess.AS"],"title_canon_sha256":"b887a5bccc360d5021f4a3de1699c6122c24c2795c00339f9f2ff8de694aec4a","abstract_canon_sha256":"8364ea58fc0ced079b439356b697f0c5c7033989935b55652f7e438b17be49fe"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:08:25.731896Z","signature_b64":"5gP1zUreqIIgXfoDxV7DUVeCbWd+Vt56nkTfynNExwthr4mLtsoUUZEO+4sD5jzry6vaXCWQ3HHp4W5qwYASBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b162e4e5069e2c768ec5b29255c2a9958937636d1071d8cf78b4ff8fc4a2b956","last_reissued_at":"2026-07-05T11:08:25.731417Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:08:25.731417Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LLM-based Generative Error Correction for Rare Words with Synthetic Data and Phonetic Context","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","eess.AS"],"primary_cat":"cs.SD","authors_text":"Hiroaki Kokubo, Masaaki Yamamoto, Natsuo Yamashita, Yohei Kawaguchi","submitted_at":"2025-05-23T02:54:52Z","abstract_excerpt":"Generative error correction (GER) with large language models (LLMs) has emerged as an effective post-processing approach to improve automatic speech recognition (ASR) performance. However, it often struggles with rare or domain-specific words due to limited training data. Furthermore, existing LLM-based GER approaches primarily rely on textual information, neglecting phonetic cues, which leads to over-correction. To address these issues, we propose a novel LLM-based GER approach that targets rare words and incorporates phonetic information. First, we generate synthetic data to contain rare wor"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.17410","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/2505.17410/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":"2505.17410","created_at":"2026-07-05T11:08:25.731476+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.17410v1","created_at":"2026-07-05T11:08:25.731476+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.17410","created_at":"2026-07-05T11:08:25.731476+00:00"},{"alias_kind":"pith_short_12","alias_value":"WFROJZIGTYWH","created_at":"2026-07-05T11:08:25.731476+00:00"},{"alias_kind":"pith_short_16","alias_value":"WFROJZIGTYWHNDWF","created_at":"2026-07-05T11:08:25.731476+00:00"},{"alias_kind":"pith_short_8","alias_value":"WFROJZIG","created_at":"2026-07-05T11:08:25.731476+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WFROJZIGTYWHNDWFWKJFLQVJSW","json":"https://pith.science/pith/WFROJZIGTYWHNDWFWKJFLQVJSW.json","graph_json":"https://pith.science/api/pith-number/WFROJZIGTYWHNDWFWKJFLQVJSW/graph.json","events_json":"https://pith.science/api/pith-number/WFROJZIGTYWHNDWFWKJFLQVJSW/events.json","paper":"https://pith.science/paper/WFROJZIG"},"agent_actions":{"view_html":"https://pith.science/pith/WFROJZIGTYWHNDWFWKJFLQVJSW","download_json":"https://pith.science/pith/WFROJZIGTYWHNDWFWKJFLQVJSW.json","view_paper":"https://pith.science/paper/WFROJZIG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.17410&json=true","fetch_graph":"https://pith.science/api/pith-number/WFROJZIGTYWHNDWFWKJFLQVJSW/graph.json","fetch_events":"https://pith.science/api/pith-number/WFROJZIGTYWHNDWFWKJFLQVJSW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WFROJZIGTYWHNDWFWKJFLQVJSW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WFROJZIGTYWHNDWFWKJFLQVJSW/action/storage_attestation","attest_author":"https://pith.science/pith/WFROJZIGTYWHNDWFWKJFLQVJSW/action/author_attestation","sign_citation":"https://pith.science/pith/WFROJZIGTYWHNDWFWKJFLQVJSW/action/citation_signature","submit_replication":"https://pith.science/pith/WFROJZIGTYWHNDWFWKJFLQVJSW/action/replication_record"}},"created_at":"2026-07-05T11:08:25.731476+00:00","updated_at":"2026-07-05T11:08:25.731476+00:00"}