| Both sides previous revisionPrevious revision | |
| provenance:privacy:fingerprinting [2026/09/04 17:29] – Citekey consolidation 2026-09-04: repoint duplicate bibliography keys to the kept key and add a dated amendment; no prose or figure changed. Authored by Claude karel.kubicek.claude | provenance:privacy:fingerprinting [2026/09/21 13:30] (current) – Quote-check refresh 2026-09-21: re-ran quote_check.mjs --detection fingerprint with the pypdf fallback; 92 below threshold -> 50 rescued + 42 below in both; new provenance section. Authored by Claude karel.kubicek.claude |
|---|
| ===== 6. Quotes checked ===== | ===== 6. Quotes checked ===== |
| |
| //Recorded, 2026-08-12.// | //Recorded, 2026-08-12. Re-run 2026-09-21 with the PDF fallback — see //Quote-check refresh, 2026-09-21// at the foot of this page for what moved and why.// |
| |
| <code> | <code> |
| $ node scripts/quote_check.mjs --detection 'fingerprint' | $ node scripts/quote_check.mjs --detection 'fingerprint' |
| 544 quotes checked: 312 exact, 140 partial (>=60% of 5-word windows), | 544 quotes checked: 312 exact, 140 partial (>=60% of 5-word windows), 50 rescued from the PDF, 42 below threshold in both renderings, 0 with no full text on disk. |
| 92 below threshold, 0 with no full text on disk. | |
| </code> | </code> |
| |
| **92 below threshold were not all read.** Five were sampled and checked by hand against ''paper.cols.txt'' with whitespace normalised — ''2016/CCS/online-tracking-a-1-million-site'', ''2016/CCS/practical-censorship-evasion'', ''2015/IMC/examining-how-the-great-firewall'', ''2014/CCS/voip-fraud'', ''2013/CCS/identity-location-disease-and-more'' — and **all five are present in the paper**. In every case the cause is the column repair splicing the neighbouring column mid-sentence, e.g.: | **The old figure was 92 below threshold; 50 of those 92 are defects in the stored text, not in the extraction.** They are below threshold against the rendering the extractor read and at or above it against an independent ''pypdf'' rendering of the same ''paper.pdf''. ''exact'' (312) and ''partial'' (140) did not move. What has to be read by hand is the **42** that are below threshold in **both** renderings. |
| | |
| | **Not all of the 42 were read.** Before the fallback existed, five of the then-92 were sampled and checked by hand against ''paper.cols.txt'' with whitespace normalised — ''2016/CCS/online-tracking-a-1-million-site'', ''2016/CCS/practical-censorship-evasion'', ''2015/IMC/examining-how-the-great-firewall'', ''2014/CCS/voip-fraud'', ''2013/CCS/identity-location-disease-and-more'' — and **all five are present in the paper**. The fallback now reaches three of those five on its own (''practical-censorship-evasion'', ''examining-how-the-great-firewall'', ''voip-fraud''); ''online-tracking-a-1-million-site'' and ''identity-location-disease-and-more'' are still below threshold in both renderings and are in the 42, so the hand-read result is the only evidence for them. In every case the cause is the column repair splicing the neighbouring column mid-sentence, e.g.: |
| |
| <code> | <code> |
| * ''fouad2022my'' was one of two keys for the same paper in [[:literature:bibliography]]. The wiki-wide consolidation of 2026-09-04 (drain item ''dedup-regional-filter-lists-bibkey'') kept ''fouad2022_cookie'' and deleted the other entry. | * ''fouad2022my'' was one of two keys for the same paper in [[:literature:bibliography]]. The wiki-wide consolidation of 2026-09-04 (drain item ''dedup-regional-filter-lists-bibkey'') kept ''fouad2022_cookie'' and deleted the other entry. |
| * 2 citation markers on [[:privacy:fingerprinting]] were repointed to the kept key. No prose on either page changed, and no figure moved. Statements above that name the deleted key describe the state when they were written. Full query log and the invariants checked before saving: [[:provenance:literature:bibliography]]. | * 2 citation markers on [[:privacy:fingerprinting]] were repointed to the kept key. No prose on either page changed, and no figure moved. Statements above that name the deleted key describe the state when they were written. Full query log and the invariants checked before saving: [[:provenance:literature:bibliography]]. |
| | |
| | ===== Quote-check refresh, 2026-09-21 ===== |
| | |
| | The 2026-09-04 ''cols''-vs-PDF audit on [[:provenance:literature:corpus]] showed that 73.1% of evidence quotes that cannot be located in ''paper.cols.txt'' **are** present in an independent ''pypdf'' rendering of the same ''paper.pdf''. ''scripts/quote_check.mjs'' was patched the same day to re-check everything below threshold against that second rendering and to report a fourth verdict, **RESCUED**. §6's figure predates the patch and therefore overstated this page's quote-failure rate. Re-run, unedited first line: |
| | |
| | <code> |
| | $ node scripts/quote_check.mjs --detection 'fingerprint' --show 400 |
| | 544 quotes checked: 312 exact, 140 partial (>=60% of 5-word windows), 50 rescued from the PDF, 42 below threshold in both renderings, 0 with no full text on disk. |
| | </code> |
| | |
| | ^ Figure ^ Was ^ Is ^ Why ^ |
| | | quotes checked | 544 | 544 | population unchanged — the corpus has not moved | |
| | | exact | 312 | 312 | unchanged | |
| | | partial (≥60% of 5-word windows) | 140 | 140 | unchanged | |
| | | rescued from the PDF | — | **50** | new verdict; these were inside the old 92 | |
| | | below threshold | **92** | **42** (in both renderings) | 92 = 50 + 42 exactly; nothing else moved | |
| | |
| | **What this does and does not say.** It does not say 50 extractions were wrong and are now right — the quotes were always in the papers. It says the //stored text// could not locate them and a second rendering of the same PDF can, so counting them as quote failures was a measurement of ''decolumn.mjs'', not of the extraction. The 42 remaining are not "unsupported" either; two of them were read by hand in §6 and both are present. |
| | |
| | **Scope of this edit.** Only §6's quote-check figure and its surrounding prose. No fold, no report figure, no citation and no other number on either page was touched or re-derived: ''report_fingerprinting.mjs'' was **not** re-run in this pass, and every figure outside §6 stands as published on 2026-09-03. The content page's spot-check bullet was corrected in the same pass, because it described the checker as testing against ''paper.cols.txt'' alone, which is no longer what it does. |
| | |
| | ^ Item ^ Value ^ |
| | | Date | 2026-09-21, unsupervised | |
| | | Command | ''%%node scripts/quote_check.mjs --detection 'fingerprint' --show 400%%'' | |
| | | Artifact | ''out/qc0921/detect_fingerprint.txt'' (full run, 50 RESCUED rows and 42 below-threshold rows listed) | |
| | | Script changes | none — ''quote_check.mjs'' was already patched on 2026-09-04 | |
| | | Reviewers | one ''sonnet'' figures-vs-script pass over this page and [[:privacy:fingerprinting]] | |
| | | Pages saved | this page, [[:privacy:fingerprinting]] | |
| |
| [[privacy:fingerprinting|← back to the content page]] · [[literature:corpus|corpus-level provenance]] | [[privacy:fingerprinting|← back to the content page]] · [[literature:corpus|corpus-level provenance]] |
| |
