Applying NLP to Support Legal Decision-making in Administrative Appeal Boards in the EU
Bibliographic record
Abstract
While Natural Language Processing (NLP) is being applied in an increasing number of contexts, including law, it remains a difficult task to leverage NLP for the purpose of real-life support of legal decision-making. This is because 1) legal-decision making must be made in a way that is sensitive not only to legislation but also to evolving case practice (prior decision-making that functions as precedent), 2) legal-decision making is sensitive to open-ended legislative language and shifting factual contexts, 3) traditional methods of NLP are capable of processing long texts, but they are suboptimal compared to novel methods, i.e., transformer-based models, e.g., BERT [1], etc. 4) however the transformer-based models are limited by maximum input lengths, which makes it difficult to apply in real-life scenarios, where legal documents exceed the maximum input length. In this paper, we show how we tackle the problem of providing NLP-based intelligence support to legal decision-makers in a real-world setting using transformer-based NLP.
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 | 0.009 | 0.049 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.003 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.005 | 0.008 |
| Open science | 0.001 | 0.004 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.006 | 0.002 |
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".