{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:B72EDDAIKJUOLIWRBMB7GDXALM","short_pith_number":"pith:B72EDDAI","schema_version":"1.0","canonical_sha256":"0ff4418c085268e5a2d10b03f30ee05b2ffe7c5dde9c42b7161de564bc521f93","source":{"kind":"arxiv","id":"1804.05958","version":1},"attestation_state":"computed","paper":{"title":"Can Neural Machine Translation be Improved with User Feedback?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.CL","authors_text":"Evgeny Matusov, Julia Kreutzer, Shahram Khadivi, Stefan Riezler","submitted_at":"2018-04-16T21:55:45Z","abstract_excerpt":"We present the first real-world application of methods for improving neural machine translation (NMT) with human reinforcement, based on explicit and implicit user feedback collected on the eBay e-commerce platform. Previous work has been confined to simulation experiments, whereas in this paper we work with real logged feedback for offline bandit learning of NMT parameters. We conduct a thorough analysis of the available explicit user judgments---five-star ratings of translation quality---and show that they are not reliable enough to yield significant improvements in bandit learning. In contr"},"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":"1804.05958","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-04-16T21:55:45Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"a51613cb7469ee617dc2c0e54afb87a08858e2a07b595c41bdce20e134f06b18","abstract_canon_sha256":"0e1b8d32352a9816d49a65076a91cdeffbaa16b6de546babfd32eb08468189f3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:18:22.230905Z","signature_b64":"9LTEtOX+rV8UORnaODoDIov8dM3IwHtC5o8r/hJDDq/xU2r7t6zljOQLhf71DgGWNp7eiFr40QQnjFqxCsEPAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0ff4418c085268e5a2d10b03f30ee05b2ffe7c5dde9c42b7161de564bc521f93","last_reissued_at":"2026-05-18T00:18:22.230188Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:18:22.230188Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Can Neural Machine Translation be Improved with User Feedback?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.CL","authors_text":"Evgeny Matusov, Julia Kreutzer, Shahram Khadivi, Stefan Riezler","submitted_at":"2018-04-16T21:55:45Z","abstract_excerpt":"We present the first real-world application of methods for improving neural machine translation (NMT) with human reinforcement, based on explicit and implicit user feedback collected on the eBay e-commerce platform. Previous work has been confined to simulation experiments, whereas in this paper we work with real logged feedback for offline bandit learning of NMT parameters. We conduct a thorough analysis of the available explicit user judgments---five-star ratings of translation quality---and show that they are not reliable enough to yield significant improvements in bandit learning. In contr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1804.05958","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":""},"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":"1804.05958","created_at":"2026-05-18T00:18:22.230289+00:00"},{"alias_kind":"arxiv_version","alias_value":"1804.05958v1","created_at":"2026-05-18T00:18:22.230289+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1804.05958","created_at":"2026-05-18T00:18:22.230289+00:00"},{"alias_kind":"pith_short_12","alias_value":"B72EDDAIKJUO","created_at":"2026-05-18T12:32:13.499390+00:00"},{"alias_kind":"pith_short_16","alias_value":"B72EDDAIKJUOLIWR","created_at":"2026-05-18T12:32:13.499390+00:00"},{"alias_kind":"pith_short_8","alias_value":"B72EDDAI","created_at":"2026-05-18T12:32:13.499390+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2009.01325","citing_title":"Learning to summarize from human feedback","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"2302.12192","citing_title":"Aligning Text-to-Image Models using Human Feedback","ref_index":8,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/B72EDDAIKJUOLIWRBMB7GDXALM","json":"https://pith.science/pith/B72EDDAIKJUOLIWRBMB7GDXALM.json","graph_json":"https://pith.science/api/pith-number/B72EDDAIKJUOLIWRBMB7GDXALM/graph.json","events_json":"https://pith.science/api/pith-number/B72EDDAIKJUOLIWRBMB7GDXALM/events.json","paper":"https://pith.science/paper/B72EDDAI"},"agent_actions":{"view_html":"https://pith.science/pith/B72EDDAIKJUOLIWRBMB7GDXALM","download_json":"https://pith.science/pith/B72EDDAIKJUOLIWRBMB7GDXALM.json","view_paper":"https://pith.science/paper/B72EDDAI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1804.05958&json=true","fetch_graph":"https://pith.science/api/pith-number/B72EDDAIKJUOLIWRBMB7GDXALM/graph.json","fetch_events":"https://pith.science/api/pith-number/B72EDDAIKJUOLIWRBMB7GDXALM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/B72EDDAIKJUOLIWRBMB7GDXALM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/B72EDDAIKJUOLIWRBMB7GDXALM/action/storage_attestation","attest_author":"https://pith.science/pith/B72EDDAIKJUOLIWRBMB7GDXALM/action/author_attestation","sign_citation":"https://pith.science/pith/B72EDDAIKJUOLIWRBMB7GDXALM/action/citation_signature","submit_replication":"https://pith.science/pith/B72EDDAIKJUOLIWRBMB7GDXALM/action/replication_record"}},"created_at":"2026-05-18T00:18:22.230289+00:00","updated_at":"2026-05-18T00:18:22.230289+00:00"}