This is an official ODERSA website. Here’s how you know

The official domain

The address of this site ends in odersa.org. Every service the association runs sits on a subdomain of odersa.org and nowhere else. If the address in your browser’s address bar does not end in odersa.org, this site is not ours.

Free, and no account

Everything is open straight away. No sign-up, no account, no password, no subscription, no advertising. Nothing is held back for those who pay, because there is nothing to pay for.

No data collected

This site does not follow you: no tracker, no tracking cookie, no measurement tool built into these pages, and nothing measured on your device. Our host counts requests in aggregate, as any server that answers does: a total, never a profile. You do not have to take our word for it: open your browser’s developer tools, go to the Network tab, and reload the page. You will see the full list of what the site asks for. Everything comes from odersa.org, nothing goes anywhere else.

Free to reuse

The content is published under the CC BY 4.0 licence. You may copy it, translate it, print it and pass it on, for your classes as much as for the people around you, on one condition only: credit ODERSA.

The register

Projects

This page indexes the division’s projects, and it does not inflate. The register counts one active project today: a system for verifiable, massively multilingual natural-language generation, which its program sets out across four axes.

Project 01 · Active

Verifiable natural-language generation

Factual documents produced by a deterministic procedure, re-verified by machine against their input facts rather than judged probable, across 1,000 equally treated languages. The project aims at guarantee rather than likelihood, and accepts what that costs: a delimited factual domain, short texts, and silence wherever an approximation would have been easy.

  • The project file – What the system produces, the three mechanically verified guarantees, the coverage sweep and the acknowledged limits.

Project identity

Status: active

Four research axes · eight properties on the register (P1-P8) · latest measurement campaign: · content under the CC BY 4.0 licence.

Formal results and public demonstration corpus: in preparation.

The project in figures

Measured on , regenerable in one command

The values below are read on the current state of the system at the displayed date. Their method, their figures and their limits are set out on the results page.

The project in figures,
Reading Value
Languages processed, equal treatment1,000
Structured fact sheets, 8 domains3,444
Cross-language conflicts, out of 1,646 compared values0
Artificial corruptions detected – 0 false alarms on 113 untouched texts105 / 105
Identical fingerprints across two separate processes147 / 147
Writing systems detected in the outputs25

The folder

The pages of the project

Each of the division’s projects lives in its own folder. This one reads in four parts, in any order.

  • Research Program – Four axes: verifiable generation, linguistic equity at scale, measured readability, bounded tutoring. Open questions and cited literature.
  • Results and Measurements – The readings of : coverage, equivalence, determinism, verifier calibration – figures, tables and threats to validity.
  • Validation Approach – The register of properties P1-P8, the negative controls, the independent external references and the scientific position.
  • Demonstrations – Dated, unretouched outputs on one subject, in twelve languages and six writing systems.

Scientific position

Often right is not guaranteed

Between a generator that is right most of the time and a system whose every document is checked against its input facts, the difference is one of nature, not of degree. This position, together with its references, is set out in full on the Validation page.

Keyboard shortcuts

TabMove from link to link
EnterOpen the link, or expand and collapse a question
?Open this help
EscClose the panel or this help