What every piece of work must meet.
Primary and secondary sourcing
We combine institutional data, public records and original data collection where regional datasets are incomplete or fragmented.
Documented provenance
Every value in a dataset can be traced to its source, and every transformation is logged.
Internal review
Publications move from Working Draft to In Review to Released only after a documented internal review of method and data.
Transparent methodology notes
Every model and dataset ships with a note on its assumptions, sources and limitations.
Rigor before productization
A finding does not become a product until its methodology has been tested against real institutional data.
Corrections
Errors are corrected openly. Corrected publications and models carry a new version and a note describing the change.
Open, Registered and Licensed.
What is public and what is shared by agreement.
Free to read
Abstracts, key points, research notes, model cards and methodology summaries.
By agreement
Model implementations, APIs and full datasets for partners and licensees — discuss licensing.
How Algocence work is numbered.
Identifiers keep every output traceable from research question to deployed system.
| Item | Format | Example |
|---|---|---|
| Research project | ALG-R-YYYY-NNN | ALG-R-2026-001 |
| Publication | Series YYYY-NN | Working Paper 2026-01 |
| Model | ALG-FAMILY-NNN | ALG-CREDIT-001 |
| Dataset | DS-NNN | DS-001 |
Citing Algocence research.
Each publication page includes a ready-to-copy citation.