Génération de base de connaissance à partir de données hétérogènes dans le monde culturel
Bibliographic record
Abstract
RÉSUMÉ: RÉSUMÉ : Le monde culturel québécois est riche et varié, et ceci se concrétise par l'importante quantité de métadonnées sur les mondes du livre et du cinéma que les acteurs gouvernementaux, aca- démiques et commerciaux ont accumulé. Cependant, ces données sont actuellement en bonne partie indisponibles au public, et sont encodées dans des bases de données dont les modèles, parfois complexes et généralement incompatibles d'une institution à l'autre, rendent l'exploi- tation difficile. De plus, sauf certaines exceptions, elles ne sont pas reliées aux métadonnées diffusées librement ailleurs sur le web, que ce soit par l'entremise de projets collaboratifs publics tels que Wikidata, ou par des acteurs tels que certaines bibliothèques nationales européennes. La création de bases de connaissances sous forme de graphes peut permettre la démocratisa- tion de ces métadonnées, en simplifiant leur exploitation et en les liant vers d'autres bases de connaissances existantes. Ce mémoire résume notre travail de création de bases de connais- sances pour les mondes du cinéma et de la littérature québécois, en particulier la modélisation de modèles ontologiques et la population des graphes à partir de sources relationnelles. Nous présentons d'abord une base de connaissances pour le domaine du cinéma québécois, qui utilise un jeu de métadonnées fourni par la Cinémathèque québécoise. À partir de scénarios d'utilisation fournis par des experts du milieu, nous développons un modèle ontologique pour ce domaine, et décrivons la conversion des données sources de leur format original vers la base de connaissances finale. ABSTRACT: Quebec's cultural world is rich and full of variety, as is illustrated through the imposing amount of cultural heritage metadata that exists. Governmental, academic and commer- cial players have accumulated a large amount of data relating to literary and film works. However, this data is currently largely unavailable to the greater public, and are held in datastores whose underlying datamodels, which are often complex and incompatible between institutions, complicate their use. On top of this, except for rare exceptions, this data is not interlinked with other linked open data sources available elsewhere on the web, whether they be public collaborative projects such as Wikidata or knowledge bases published by national librairies. The development of knowledge bases in graph form can aid in democratising this metadata, by simplifying its exploitation and allowing it to be linked with existing, open knowledge bases. This memoir summarizes our work, which is the creation of knowledge bases for Quebec's cinema and literature data. In particular, our work focuses on modelling and populating such knowledge bases from existing relational databases. We first model a knowledge base for Quebec's film world, which uses a dataset provided by the Cinémathèque québécoise. Our use cases, provided by domain experts, guide our development of an ontological model for this domain. We describe the translation of this source data from its original format towards the final knowledge base.
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 | 0.004 | 0.021 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.005 | 0.005 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.006 | 0.006 |
| Open science | 0.001 | 0.005 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.009 | 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".