This index contains scientific papers and publications related to SHARE (Survey of Health, Ageing and Retirement in Europe). Search using natural language queries to find relevant research on topics like health, aging, retirement, cognition, and socioeconomic factors in European populations.
SARA SHARE Assisted Research Agent
Local LLM
SADA SHARE Assisted Data Analyst
SHARE MicrodataMCP / AI Service Access and Use Conditions (SHARE-ERIC)
Access to the MCP server and associated AI services is provided exclusively to registered SHARE-ERIC data users in good standing.
Use of this service and any AI outputs is subject to and governed by the SHARE-ERIC Conditions of Use and is strictly limited to scientific research purposes consistent with each user's registered SHARE project and user licence.
No additional rights are granted beyond those already conferred under the SHARE data access framework.
Note on Scientific Research Use
Under EU copyright law, in particular Article 3 of Directive (EU) 2019/790, research organisations are permitted to carry out text and data mining (TDM) of works to which they have lawful access for the purposes of scientific research.
The MCP / AI service forms part of SHARE-ERIC's controlled research infrastructure. Access is therefore limited to registered SHARE users and may only be used for scientific research in accordance with the SHARE Conditions of Use.
Use of the MCP / AI service for commercial purposes, general public services, or other non-research activities is not permitted.
Usage Metrics and Data Collection
To operate and improve service reliability, SHARE-ERIC collects privacy-safe, aggregated usage metrics for the research portal and AI endpoints.
Collected metrics include:
- Service and endpoint name
- Time bucket (hour/day) and total request counts
- HTTP status classes (2xx/4xx/5xx) and latency percentiles
- For search endpoints only: query-length buckets and result-count buckets
Not collected in these metrics:
- Raw query text, prompts, or model responses
- Uploaded document contents
- API keys or authentication secrets
- Full IP addresses or direct personal identifiers
Public metrics views apply low-volume suppression (k-anonymity threshold) to reduce re-identification risk.