Building a searchable online corpus of Australian and New Zealand aligned speech
Bibliographic record
Abstract
Advances in automatic speech recognition technology, increases in bandwidth availability, and the widespread use of video streaming and sharing platforms have opened new horizons for corpus phonetics. CoANZSE Audio, a searchable online version of the Corpus of Australian and New Zealand Spoken English, provides access to over 195 million words of transcribed speech from transcripts of videos uploaded to YouTube by councils and other local government entities in Australia and New Zealand. Audio and forced alignment files are also available, making the resource suitable for the investigation of a range of research questions pertaining to morphosyntax, phonetics, and discourse. The resource, which is freely available via login through CLARIN, Europe’s main language resources infrastructure network, was created through the use of open-source tools and software: yt-dlp, a Python library for collecting data from video and streaming websites; the Montreal Forced Aligner, a recent neural network alignment suite; and Parselmouth-Praat, Python bindings for the Praat acoustic analysis software. The website is powered by BlackLab, which combines a powerful search engine based on Apache Lucene with an intuitive web frontend. CoANZSE Audio may be useful for the investigation of regional differentiation of language features, and with additional annotation, differences in feature use according to social or demographic groups. Recent applications have included studies of double modals, a rare syntactic feature, and apology sequences. The nature of the audio and alignment data may make the resource especially suitable for the study of regional phonetic variation. Furthermore, the methods used to create the resource may be of interest to researchers seeking to adopt a pipeline approach for the creation of specialized corpora from publicly available online content.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.014 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.007 | 0.008 |
| Science and technology studies | 0.003 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.002 | 0.005 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.028 | 0.015 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".