Fix Linear B Wiktionary tr= extraction bug causing 5 wrong IPA values
Alvin
105ea5a
Root cause: re.search(r"|tr=([^|}]+)") matched the first tr= in the
entire wikitext, which for entries with {{quote-book}} templates was a
full tablet transliteration (e.g., "o-di-do-si du-ru-to-mo / ..."),
not the headword. This caused 5 entries (i-qo, ma-ra-tu-wo, o-do-ke,
o-di-do-si, wo-ro-ki-jo-ne-jo) to receive IPA from wrong Wiktionary
pages (po-ro, sa-sa-ma, a-re-pa, a-ko-so-ne, e-re-mo).
Fix: prefer unicode_to_translit(title) as primary transliteration,
only fall back to tr= from {{head|gmy|...}} templates. Removes 6
spurious entries created by wrong tr= extraction. Count: 2484 → 2478.
Add Linear B (Mycenaean Greek) dataset
Alvin
75d5ab7
Build structured Linear B dataset from 4 CC-BY-SA compatible sources:
- Unicode UCD: 211 signs (88 syllabograms + 123 ideograms)
- jhnwnstd/shannon: 2,272 words from Linear B Lexicon (MIT)
- Wiktionary gmy: 170 words with 46 expert IPA transcriptions
- IE-CoR: 42 words with expert IPA and Concepticon IDs
Output: 2,484 words with IPA, SCA encoding, glosses, source
attribution. Sign inventory with 74 Ventris grid IPA mappings.
IPA mapping follows Ventris & Chadwick (1973) and Hooker (1980).
Adversarial audit passed all 6 checks.
Add Linear A Phonotactics Validation dataset
Alvin
f304b7a
1,414,143 gold-standard cognate pairs between 360 languages that
phonotactically resemble Linear A (CV open syllables, no clusters,
final vowels). Every pair has Confidence=certain and Phono_Quality=strong
(SCA Score >= 0.5). Threshold: Linear_A_Score >= 0.80 for both languages.
Includes typological analysis script and full 1,047-language ranking.
- Added 15th column 'Phono_Quality' (strong/moderate/weak/none/unscored)
based on SCA score thresholds to flag phonologically divergent cognates
- Downgraded 27,059 Robbeets cross-family pairs from certain to contested
- Fixed 2,906 Sino-Tibetan pairs from certain to doubtful (STEDT doubt markers)
- Rebuilt Parquet (31.3 MB) with new schema
- Added changelog entry 007 and flagging script
31 new ancient/historical languages: Luwian, Middle Persian, Tocharian A/B,
Mycenaean Greek, Classical Armenian, Gaulish, Oscan, Umbrian, Bactrian,
Sogdian, Parthian, Khotanese, Khwarazmian, Old Japanese, Old Turkic, etc.
83 new languages total. 4-agent adversarial audit: code integrity PASS,
20/20 random traceback PASS, academic credentials CONDITIONAL PASS
(Kitchen Semitic excluded for CC-BY-NC license), data quality PASS after
IPA resolution fix. See docs/changelog/006_tier1_cldf_ingestion.md.
Add raw scraped sources, staging intermediates, and missing audit trail
Alvin
3347240
Provenance chain for full traceability: raw scraped HTML/XML/JSON/CSV
from all ingest scripts, intermediate staging cognate pair TSVs, and
xsa_audit.json. Enables debugging data quality issues back to source.
Add complete dataset: all sources, metadata, scripts, docs, and phylo enrichment
Alvin
26786e3
Syncs HuggingFace repo with full GitHub dataset:
- 1,177 lexicons (modern + ancient languages)
- 1,110 language profiles
- 102 audit trails
- 53 validation sets
- cognate_pipeline package (full Python source)
- All scripts (ingest, build, validate, scrape)
- All docs (PRDs, adversarial audits, database reference)
- Phylogenetic enrichment: phylo_pairs.tsv (386K pairs), Glottolog CLDF tree
- Additional data: gothic, iberian, ugaritic, religious_terms, linear_a, cited_sources
