Extraction of nominative entities, an opportunity for the cultural sector ?
Bibliographic record
Abstract
[etude] Les champs de metadonnees non structures tels que « description » offrent une plus-value considerable a la comprehension pour les utilisateurs finals. Neanmoins, leur caractere non structure les rend peu exploitables dans un contexte electronique et d’automatisation. Cet article explore les possibilites et les limitations de la reconnaissance d’entites nommees (« Named-Entity Recognition », NER) et de l’extraction terminologique (« Term Extraction », TE) dans la prospection de donnees non structurees afin d’en extraire des concepts significatifs. Ces concepts permettent de tirer parti d’une recherche et d’une navigation ameliorees, mais peuvent egalement jouer un role tres important dans la recherche en humanites numeriques. A travers une etude de cas basee sur les champs de description des archives historiques de la ville de Quebec, les auteurs, Simon HENGCHEN, Seth van HOOLAND, Ruben VERBORGH et Max DE WILDE, proposent une evaluation de quatre services tiers d’extraction d’entites afin de promouvoir l’experimentation de la reconnaissance d’entites nommees et l’extraction terminologique. Dans le but de couvrir autant le NER que la TE, ils utilisent, pour l’evaluation des entites nommees, une approche quantitative basee sur la precision, le rappel et le F-score calcules sur la base d’un referent manuel (« gold standard corpus »). Une seconde approche, plus qualitative, permet ensuite de prendre en compte la pertinence des termes extraits et aborde la question du multilinguisme.
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 | 0.005 | 0.015 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.004 | 0.007 |
| Science and technology studies | 0.003 | 0.002 |
| Scholarly communication | 0.006 | 0.010 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.008 | 0.003 |
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".