CustomGPT · UNESCO Datahub SDG alignment + domain classification

Research alignment

A research-focused CustomGPT connected with UNESCO Datahub to map Sustainable Development Goals (SDGs), surface domain knowledge, and draft briefs.
From prompt → structured evidence
Ask for SDG mappings, dataset-backed summaries, or “what to measure” checklists. The model retrieves UNESCO records and shapes them into compact research scaffolds you can adapt to proposals, evaluation plans, or exploratory studies.
Open Research Alignment in ChatGPT Read privacy notice
SDGs
Maps prompts to SDG targets and indicators (as a working classification, not a verdict).
Domains
Suggests disciplines, subfields, and related literatures to widen the query space.
Evidence
Pulls public UNESCO Datahub records and turns them into citations-ready notes.

What it’s good at

Designed for research scaffolding, not final authority.

SDG alignment

Turns messy prompts into SDG-relevant framings, with suggested indicators and dataset pointers.

Domain mapping

Creates a “discipline graph” around your topic: fields, subfields, likely methods, and related terms.

Dataset discovery

Finds UNESCO Datahub datasets and proposes how they can support research questions and evaluation rubrics.

How it works

High-level model of the pipeline.
1

You write a prompt describing a paper, a claim, or a research goal (e.g., “classify this abstract to SDGs”).

2

The GPT forms a query and calls the UNESCO Datahub API endpoints (public datasets) to retrieve relevant records.

3

You get a structured output: SDG mapping, domain taxonomies, dataset references, and next-step research prompts.

Disclaimer: Research alignment is a tool for ideation and synthesis. It can be wrong, outdated, or overconfident.

Privacy notice

A privacy notice suitable for a GPT “Actions” privacy policy URL.
Controller: Rafail Konstantinidis Contact: rafailkonstantinidisx@gmail.com Last updated: 2026-01-15
What data is processed?
  • What you type into ChatGPT (prompts, uploaded text you paste, and any details you include) may be processed by OpenAI to provide the chat experience.
  • When the UNESCO action is used, query parameters derived from your prompt are sent to UNESCO Datahub API endpoints to retrieve public dataset records.
  • If you contact me directly (email), I can respond with relevant information about The Concept (how it works, knowledge sources, and usage instructions). Email messages are separate from your ChatGPT conversations: I do not receive or have access to your ChatGPT chat history through email. Your use of OpenAI/ChatGPT and UNESCO Datahub is governed by their own terms and privacy policies, which you should review independently.
Tip: Avoid entering sensitive personal data (health, IDs, passwords) into prompts.
Third parties involved
  • OpenAI / ChatGPT processes your conversation to run the GPT.
  • UNESCO Datahub receives API requests to return public data records.
Important: Third parties may collect standard technical metadata (e.g., request logs) according to their own policies.
Do you store personal data?

This website page does not require you to submit personal data to read it. For the GPT action itself: I do not intentionally request personal data. If any server-side logging is enabled for debugging (e.g., error logs), it should be limited, protected, and retained for a short period.

  • Goal: operate the tool and troubleshoot issues.
  • Retention: if logs exist, a reasonable default is 30 days (or less) unless security/abuse investigations require longer.
Legal basis & user rights (GDPR-friendly)
  • Legal basis: legitimate interests (tool operation) and/or consent (if you email or opt into optional analytics/cookies).
  • Your rights: access, rectification, deletion, restriction, objection, and portability where applicable.
  • How to exercise rights: email me at rafailkonstantinidisx@gmail.com with “The Concept — Privacy” in the subject.
Security & best practices
  • Do not share secrets (API keys, passwords) in prompts.
  • Prefer abstractions: use “a user” / “a client” instead of identifiable names.
  • If you believe a privacy issue occurred, contact me immediately.
Non-legal note: This text is a practical, plain-language notice for a small creator project. If you’re publishing at scale, a lawyer-reviewed privacy policy is the correct move.