Bibliographic record
Abstract
ICAME conferences are invariably innovatory: new descriptions from new corpora using new techniques.In this regard, ICAME19-98 was no exception, but three new dimension s made it special: widespread use of very large corpora such as the Cobuild Bank of English (BojE) , the British National Corpus (BNC), CD-ROM newspaper corpora, and even the internet (or world- wide web); as a result, the use of corpora for lexical studies; and the development of parsed corpus concordances.No doubt the debate over the relative merits of corpus size versus register balance will continue.But there is certainly merit in pursuing size if, as these papers show, it yields good results.According to Kjellmer, for instance, 'the arrival on the scene of large-scale language corpora has now made it possible for us to register even very marginal variation phenomena in a fairly systematic way, phenomena which may thus be the precursors of changes to come' (this volume: liS).The spoken component of the BofE (currently 325,000,000 words) is used in the paper by Susan Blackwell as a resource of data for discourse uses of three words: honest (on its own, but also its various collocations to be (quite) honest, let's be honest, I'll be (quite) honest, and honest to God), look (especially utterance-initially), and well when tagged as a 'formulaic interactive expression'.A very different use of the BofE features in the theoretical paper by Oliver Mason on the possibilities and techniques for measuring collocations.Mason's purpose is to demonstrate that 'the span for computing a word's collocates can be determined empirically'; to this end, he uses the BofE as a control corpus, and a sample of it for detailed investigation.The BofE is also used by Vincent Ooi to generate examples of words under investigation by him as Asian words.The occurrence of some 'eastern' words in this 'western' corpus may be explained by the BofE' s high newspaper constituency .The Cobuild Direct Corpus (CDC) (50 million words of the BofE available as a CD-ROM) is used by Goran Kjellmer in his study of complements of the lexical verb TRY.He finds that, whereas the TRY + bare infinitive neither occurs in any of the earlier corpora (the Brown Corpus, the Lancaster-Oslo/Bergen Corpus and the London-Lund Corpus (LLC) , nor in the Freiburg LOB Corpus (the 1991 equivalent of LOB), there are 47 occurrences in CDC, with most examples coming from 'Australian News' and ' UK Spoken ' sections .Interpreting the behaviour of these data using criteria for auxiliaryhood, Kjellmer conclude s that TRY is beginning to move towards auxiliaryhood status.The main alternative to the BofE is the 100,000,000 word BNe.The webreport hy Lou Burnard http://users.ox.ac.uk/-Ioulreports/9805icame.htmfinds that nearly 20 of the 70 or so papers and posters at the conference made use of BNC data, but this has primarily to do with accessibility and cost, now that the John M.
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 | 0.001 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.003 | 0.001 |
| Scholarly communication | 0.005 | 0.004 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.639 | 0.474 |
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".