Bibliographic record
Abstract
In recent years, Information Retrieval (IR) has evolved from ad hoc document retrieval to passage and answer retrieval, incorporating downstream Natural Language Processing (NLP). This led to remarkable progress in models when evaluated on early precision, yet at the same time, the potential to improve recall aspects has received less attention. This paper investigates an extremely high-recall task by a reproducibility study on a massive collection of merger and acquisition documents in due diligence passage retrieval. We have replicated previous work using Conditional Random Fields (CRF) and introduced a Python version of the effective CRFsuite approach. In addition, we explore the utility of open-source and closed-source Large Language Models (LLMs) with zero-shot and few-shot learning techniques on 50 different due diligence topics. Our findings reveal the potential for few-shot learning in due diligence, delivering acceptable levels of performance in terms of recall, marking an essential step towards developing advanced due diligence models that minimize the dependency on extensive training data typically required by domain-specific IR and NLP models. More generally, our results are an important first step toward developing advanced due diligence models for any legal information need.
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 | 0.013 | 0.073 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.003 | 0.002 |
| Scholarly communication | 0.003 | 0.008 |
| Open science | 0.003 | 0.003 |
| Research integrity | 0.002 | 0.005 |
| Insufficient payload (model declined to judge) | 0.005 | 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".