Bibliographic record
Abstract
During the rapid growth of China's commercial health insurance industry in the 2010s, many third-party administrators (TPAs) emerged to provide administrative services to primary insurers. However, most of these TPAs were small and medium-sized companies (SMEs) that offered poor-quality services and were often short-lived. This made them incompatible with the long time horizons and high risk management requirements of the insurance sector. There were also additional structural, institutional, and technical obstacles that primary insurers had to overcome to collaborate with TPAs. To address these pain points, in 2010, Steve Zhang, then Managing Director of Munich Re Life China, and Eric Zhao, general manager of the operations department responsible for insurance innovation, established HAP (Healthcare Assistance Platform) with a view to creating a one-stop solution for the customers of primary insurers by integrating top TPAs under one roof. After a decade of growth, by 2020, the company offered more than 30 services through this platform. It had also developed a set of criteria used to screen and evaluate TPAs, ensuring the quality of its services, which earned the platform recognition from primary insurers. As direct competition in the health insurance market intensified, and insurers became increasingly aware of the importance of customer service and data collection, many primary insurers started to develop their own service platforms, while reinsurers and TPAs also began to experiment in this direction. HAP was originally launched as a supporting service for Munich Re, so it wasn't placed under pressure to grow by the company. However, as HAP's operations matured and market competition intensified, Zhang began to entertain the possibility that the platform might grow its user base from three million to ten million in three to five years or perhaps even become a future profit center for Munich Re. However, he needed to carefully consider how this objective might be accomplished.
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 | 0.002 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.005 | 0.004 |
| Open science | 0.001 | 0.005 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.077 | 0.033 |
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".