home edit page issue tracker

This page pertains to UD version 2.

Metadata harmonisation: align spoken-data fields with UniDive naming conventions

Back to Khoekhoe KDT · Back to index

Repo: https://github.com/UniversalDependencies/UD_Khoekhoe-KDT

Cross-posting from the UniDive WG1 T1.5 (spoken language guidelines) metadata harmonisation review. We compared UD_Khoekhoe-KDT’s current CoNLL-U metadata against the proposed naming conventions (see also the full treebank status table). This is a suggestion for maintainers to review - feel free to push back on anything that doesn’t fit the corpus. The comparison was carried out semi-automatically with the help of Claude (Anthropic); errors or misunderstandings are possible, so please double-check anything unclear.

1. Is the spoken portion identifiable?

This treebank mixes spoken and written material but its .conllu files don’t explicitly mark which sentences are spoken. We looked for a pattern in the data (a reasonable guess):

Finding: Likely identifiable via the document_id prefix, which names the source type.

Evidence: document_id prefixes (small, clean set): book (15), grammar (2), film (2), conversation (1). conversation and film (transcribed dialogue/subtitles) are plausibly spoken; book/grammar are written.

Suggestion: Add # modality = spoken to documents whose document_id starts with conversation or film - please confirm whether film here means subtitle/transcript text.

2. Sentence-level (naming conventions)

Field Suggestion
english change to text_eng (ISO 639-3)
parallel_id corpus-specific (sentence-level) - verify against metadata.html

3. Token-level (naming conventions)

Field Suggestion
OrigLang rename to OrigLang
Lang rename to Lang

Implementation notes

Quick search & replace

Needs a small script

Needs manual input from maintainers