Skolt_Sami Giellagas
Overview
| type | mixed |
| available since | 2.5 |
| link | https://github.com/UniversalDependencies/UD_Skolt_Sami-Giellagas |
| genre | nonfiction news spoken |
| contributors | Rueter, Jack; Juutinen, Markus; Tyers, Francis; Pirinen, Tommi A; Hämäläinen, Mika |
| sentences | 261 |
| tokens | 3049 |
Issue draft: UD_Skolt_Sami-Giellagas
Modality identification
Is spoken part clearly identifiable? Yes - it aligns exactly with the file split. Per the README, the corpus “originally consists of twenty translated sentences … made by Hilkka Fofonoff from the Finnish texts”, with all subsequent sentences from the Giellagas Corpus of Spoken Saami Languages. Confirmed in the data: sms_giellagas-ud-train.conllu contains exactly those 20 sentences (sent_id prefix FofonoffHilkka_brat_2018) - written, translated from Finnish. sms_giellagas-ud-test.conllu contains the remaining 241 sentences - spoken, drawn from field recordings (sent_id prefixes like 11308_1a::<timestamp>, NA2_00635_1az::<timestamp>) and published transcriptions of recorded speech (kotus-skak2010-*). Note: a handful of test sentences (SK2020-*, SKKV2020:*) seem grammar-book examples rather than recordings.
Metadata review
languages and translation(s)
| Field | Advice |
|---|---|
text_fi |
rename to text_fin (ISO 639-3) |
text_en |
rename to text_eng (ISO 639-3) |
doc (and paragraphs) metadata
no document_id exists, but it can be derived from the source-identifier prefix of sent_id (the part before the timestamp/counter suffix), e.g. 11308_1a and NA2_00635_1az (recordings), kotus-skak2010-1-1 (published transcription), SK2020, SKKV2020, Juutinen2023, FofonoffHilkka_brat_2018 (other sources). The exact delimiter varies by source (:: before timestamps, :/- before a numeric counter otherwise) - please confirm the intended document granularity with the maintainer.
| Field | Advice |
|---|---|
| — | derive # document_id from the sent_id source-identifier prefix |
modality metadata
add # modality = written to sms_giellagas-ud-train.conllu sentences and # modality = spoken to sms_giellagas-ud-test.conllu sentences (see Modality identification above).
additional fields
aannotation (likely a typo for “annotation”) packs three XML-style attributes into one comment line: # aannotation="yes" begintime="0:39:13" endtime="0:09:18" (endtime is sometimes empty). We suggest splitting these into the standard sentence-level timing fields; sound_alignment_begin/sound_alignment_end are specified in milliseconds, so begintime/endtime (currently H:MM:SS) need converting, not just copying.
| Field | Advice |
|---|---|
aannotation (begintime) |
convert H:MM:SS to milliseconds, split into sound_alignment_begin |
aannotation (endtime) |
convert H:MM:SS to milliseconds, split into sound_alignment_end (when present); derive duration (ms) from begin/end when both are available |
text_olo |
unsure |
text_mdf |
unsure |