Amélioration continue d’une chaîne de traitement de documents avec l’apprentissage par renforcement
Bibliographic record
Abstract
Nous modelisons une chaine de traitement de documents comme un processus de deci-sion markovien, et nous utilisons l'apprentissage par renforcement afin de permettre a l'agent d'apprendre a construire des chaines adaptees a la volee, et de les ameliorer en continu. Nous construisons une plateforme qui nous permet de mesurer l'impact sur l'apprentissage de divers modeles, services web, algorithmes, parametres, etc. Nous l'appliquons dans un contexte indus-triel, specifiquement a une chaine visant a extraire des evenements dans des volumes massifs de documents provenant de pages web et d'autres sources ouvertes. Nous visons a reduire la charge des analystes humains, l'agent apprenant a ameliorer la chaine, guide par leurs retours (feedback) sur les evenements extraits. Pour ceci, nous explorons des types de retours differents, d'un feedback numerique requerant un important calibrage, a un feedback qualitatif, beaucoup plus intuitif et demandant peu, voire pas du tout, de calibrage. Nous menons des experiences, d'abord avec un feedback numerique, puis nous montrons qu'un feedback qualitatif permet tou-jours a l'agent d'apprendre efficacement. ABSTRACT. We model a document treatment chain as a Markov Decision Process, and use reinforcement learning to allow the agent to learn to construct and continuously improve custom-made chains on the fly . We build a platform which enables us to measure the impact on the. Cet article est une version etendue d'un article presente aux 26es Journees francophones d'Ingenierie des Connaissances Nicart et al. (2015).
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
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 teacher head, 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".