{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2018:U3NPTDEEN5LLIWI74NCABYDWM5","short_pith_number":"pith:U3NPTDEE","canonical_record":{"source":{"id":"1809.01448","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-09-05T11:55:05Z","cross_cats_sorted":[],"title_canon_sha256":"60d965df291b21f2c709dd6d936b27417169a99c0de2521068c0c5439f74e3b5","abstract_canon_sha256":"34d2910a37f6b500ebc0bc284eae2c9910454b37cfa67d0e5ee79c45930490fc"},"schema_version":"1.0"},"canonical_sha256":"a6daf98c846f56b4591fe34400e0766766485d0f8f8835aebcc4e26f193468d9","source":{"kind":"arxiv","id":"1809.01448","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1809.01448","created_at":"2026-05-18T00:06:27Z"},{"alias_kind":"arxiv_version","alias_value":"1809.01448v1","created_at":"2026-05-18T00:06:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1809.01448","created_at":"2026-05-18T00:06:27Z"},{"alias_kind":"pith_short_12","alias_value":"U3NPTDEEN5LL","created_at":"2026-05-18T12:32:56Z"},{"alias_kind":"pith_short_16","alias_value":"U3NPTDEEN5LLIWI7","created_at":"2026-05-18T12:32:56Z"},{"alias_kind":"pith_short_8","alias_value":"U3NPTDEE","created_at":"2026-05-18T12:32:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2018:U3NPTDEEN5LLIWI74NCABYDWM5","target":"record","payload":{"canonical_record":{"source":{"id":"1809.01448","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-09-05T11:55:05Z","cross_cats_sorted":[],"title_canon_sha256":"60d965df291b21f2c709dd6d936b27417169a99c0de2521068c0c5439f74e3b5","abstract_canon_sha256":"34d2910a37f6b500ebc0bc284eae2c9910454b37cfa67d0e5ee79c45930490fc"},"schema_version":"1.0"},"canonical_sha256":"a6daf98c846f56b4591fe34400e0766766485d0f8f8835aebcc4e26f193468d9","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:06:27.622609Z","signature_b64":"8sbI6kEiAe1ut7V0zzCDYEufBKGM//8fs0JpVu2G0zeXge996Om2aNz4hqgYedgetzaEfJqCT9VH5O8n7q0tDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a6daf98c846f56b4591fe34400e0766766485d0f8f8835aebcc4e26f193468d9","last_reissued_at":"2026-05-18T00:06:27.622031Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:06:27.622031Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1809.01448","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T00:06:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MaS7XMK0PSYdMGeH6oowyeJ64bs+PiGbqkIf4AQvlQlAh1AD7MNNacrbyZlUcWN0NZv6XMhAfhcwsvPUEMygCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-03T10:43:21.408962Z"},"content_sha256":"a2daa4637d44190bbebed4f54eb4415a13c617c46c0719f496d5b00f8a79a3f7","schema_version":"1.0","event_id":"sha256:a2daa4637d44190bbebed4f54eb4415a13c617c46c0719f496d5b00f8a79a3f7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2018:U3NPTDEEN5LLIWI74NCABYDWM5","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Appendix - Recommended Statistical Significance Tests for NLP Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Roi Reichart, Rotem Dror","submitted_at":"2018-09-05T11:55:05Z","abstract_excerpt":"Statistical significance testing plays an important role when drawing conclusions from experimental results in NLP papers. Particularly, it is a valuable tool when one would like to establish the superiority of one algorithm over another. This appendix complements the guide for testing statistical significance in NLP presented in \\cite{dror2018hitchhiker} by proposing valid statistical tests for the common tasks and evaluation measures in the field."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1809.01448","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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T00:06:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TD+w9eOuobyndFqygcGWp2Ana1K515WPQZYdM5P6A4oG2X5jaKZO+0YtGE3xOzGEMQWJOBxduU9URs8s02Y4Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-03T10:43:21.409304Z"},"content_sha256":"b49c51429aba53589b22f8ea196e6d40f6b4a18c789ff4b6c1b3e171befabfbb","schema_version":"1.0","event_id":"sha256:b49c51429aba53589b22f8ea196e6d40f6b4a18c789ff4b6c1b3e171befabfbb"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/U3NPTDEEN5LLIWI74NCABYDWM5/bundle.json","state_url":"https://pith.science/pith/U3NPTDEEN5LLIWI74NCABYDWM5/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/U3NPTDEEN5LLIWI74NCABYDWM5/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-06-03T10:43:21Z","links":{"resolver":"https://pith.science/pith/U3NPTDEEN5LLIWI74NCABYDWM5","bundle":"https://pith.science/pith/U3NPTDEEN5LLIWI74NCABYDWM5/bundle.json","state":"https://pith.science/pith/U3NPTDEEN5LLIWI74NCABYDWM5/state.json","well_known_bundle":"https://pith.science/.well-known/pith/U3NPTDEEN5LLIWI74NCABYDWM5/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2018:U3NPTDEEN5LLIWI74NCABYDWM5","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"34d2910a37f6b500ebc0bc284eae2c9910454b37cfa67d0e5ee79c45930490fc","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-09-05T11:55:05Z","title_canon_sha256":"60d965df291b21f2c709dd6d936b27417169a99c0de2521068c0c5439f74e3b5"},"schema_version":"1.0","source":{"id":"1809.01448","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1809.01448","created_at":"2026-05-18T00:06:27Z"},{"alias_kind":"arxiv_version","alias_value":"1809.01448v1","created_at":"2026-05-18T00:06:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1809.01448","created_at":"2026-05-18T00:06:27Z"},{"alias_kind":"pith_short_12","alias_value":"U3NPTDEEN5LL","created_at":"2026-05-18T12:32:56Z"},{"alias_kind":"pith_short_16","alias_value":"U3NPTDEEN5LLIWI7","created_at":"2026-05-18T12:32:56Z"},{"alias_kind":"pith_short_8","alias_value":"U3NPTDEE","created_at":"2026-05-18T12:32:56Z"}],"graph_snapshots":[{"event_id":"sha256:b49c51429aba53589b22f8ea196e6d40f6b4a18c789ff4b6c1b3e171befabfbb","target":"graph","created_at":"2026-05-18T00:06:27Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"Statistical significance testing plays an important role when drawing conclusions from experimental results in NLP papers. Particularly, it is a valuable tool when one would like to establish the superiority of one algorithm over another. This appendix complements the guide for testing statistical significance in NLP presented in \\cite{dror2018hitchhiker} by proposing valid statistical tests for the common tasks and evaluation measures in the field.","authors_text":"Roi Reichart, Rotem Dror","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-09-05T11:55:05Z","title":"Appendix - Recommended Statistical Significance Tests for NLP Tasks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1809.01448","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:a2daa4637d44190bbebed4f54eb4415a13c617c46c0719f496d5b00f8a79a3f7","target":"record","created_at":"2026-05-18T00:06:27Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"34d2910a37f6b500ebc0bc284eae2c9910454b37cfa67d0e5ee79c45930490fc","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-09-05T11:55:05Z","title_canon_sha256":"60d965df291b21f2c709dd6d936b27417169a99c0de2521068c0c5439f74e3b5"},"schema_version":"1.0","source":{"id":"1809.01448","kind":"arxiv","version":1}},"canonical_sha256":"a6daf98c846f56b4591fe34400e0766766485d0f8f8835aebcc4e26f193468d9","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a6daf98c846f56b4591fe34400e0766766485d0f8f8835aebcc4e26f193468d9","first_computed_at":"2026-05-18T00:06:27.622031Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T00:06:27.622031Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8sbI6kEiAe1ut7V0zzCDYEufBKGM//8fs0JpVu2G0zeXge996Om2aNz4hqgYedgetzaEfJqCT9VH5O8n7q0tDg==","signature_status":"signed_v1","signed_at":"2026-05-18T00:06:27.622609Z","signed_message":"canonical_sha256_bytes"},"source_id":"1809.01448","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a2daa4637d44190bbebed4f54eb4415a13c617c46c0719f496d5b00f8a79a3f7","sha256:b49c51429aba53589b22f8ea196e6d40f6b4a18c789ff4b6c1b3e171befabfbb"],"state_sha256":"ea4664b0ea79b6c457cce1d2a053781937464a1e522850c516b345f6adbdd48f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ie+40USWMqmHweDrGTJiiBGZVbEoPN22k+PfE0Ak0mtJqojsOU5jRUHTRNsFvualLd6efD0X7XvyQCMV4nMVBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-03T10:43:21.411647Z","bundle_sha256":"f93d946a56b52b87623a01fb739efc03327d7a345b956d0dced599b2be7d5710"}}