An NLP-based novel approach for assessing national influence in clause dissemination across bilateral investment treaties
Bibliographic record
Abstract
International investment agreements (IIAs) promote foreign investment. However, they can undermine crucial health programs, creating a dilemma for governments between corporate and public health interests. For this reason, including clauses that safeguard health has become an essential practice in IIAs. According to the current literature, some countries have played a pivotal role in leading this inclusion, while others follow the former ones. However, the existing literature needs a unique approach that can quantify the influence strength of a country in disseminating clauses that explicitly mention health provisions to others. Following an NLP (Natural Language Processing)-based text similarity analysis of Bilateral Investment Treaties (BITs), this study proposes a metric, 'Influence' (INF), which provides insights into the role of different countries or regions in the propagation of IIA texts among BITs. We demonstrate a comprehensive application of this metric using a large agreement dataset. Our findings from this application corroborate the evidence in the current literature, supporting the validity of the proposed metric. According to the INF, Germany, Canada, and Brazil emerged as the most influential players in defensive, neutral, and offensive health mentions, respectively. These countries wield substantial bargaining power in international investment law and policy, and their innovative approaches to BITs set a path for others to follow. These countries provide crucial insights into the direction and sources of influence of international investment regulations to safeguard health. The proposed metric holds substantial usage for policymakers and investors. This can help them identify vital global countries in IIA text dissemination and create new policy guidelines to safeguard health while balancing economic development and public health protection. A software tool based on the proposed INF measure can be found at https://inftool.com/.
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 | 0.007 | 0.041 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.025 | 0.016 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.005 | 0.007 |
| Open science | 0.002 | 0.004 |
| Research integrity | 0.003 | 0.002 |
| Insufficient payload (model declined to judge) | 0.006 | 0.004 |
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".