Third-Party Notices

The word-lookup dictionary served inside LinCaden is imported from the following public datasets. Attribution is required by their licenses and is reproduced here verbatim.

WortUniversum English Vocabulary Database — "WortUniversum English Vocabulary Database by cstr, CC BY-SA 4.0, via Hugging Face". Source: huggingface.co/datasets/cstr/grundwortschatz-voc-en — License: CC BY-SA 4.0

Kaikki/Wiktextract English Wiktionary extract — "Wiktionary contributors, extracted by wiktextract/kaikki.org, CC BY-SA 4.0 (some revisions also available under GFDL)". Source: kaikki.org/dictionary/English — License: CC BY-SA 4.0

Optional merge sources that may be imported in the future: VNEDICT English-Vietnamese dictionary by Paul Denisowski (CC BY 3.0, denisowski.org/Vietnamese/vnedict.txt) and CMUdict, the Carnegie Mellon Pronouncing Dictionary (permissive; acknowledgement of use requested, github.com/cmusphinx/cmudict).

Both CC BY-SA 4.0 datasets carry a share-alike obligation: any adapted copy of these datasets that we redistribute would itself have to be licensed under CC BY-SA 4.0. LinCaden uses the data for in-app lookup only and does not redistribute or sell the datasets — see the usage policy below.

These providers are queried on demand while you use word-lookup features. Their responses are not stored or redistributed as a dataset.

Tatoeba — example sentences are fetched from tatoeba.org under CC BY 2.0 FR. Attribution for Tatoeba content is maintained centrally on this page. Translations are requested only after you explicitly pick a target language.

Pexels — photos are fetched from pexels.com under the Pexels license. The photographer is credited and links to their Pexels profile.

Datamuse — related-word suggestions via the Datamuse API (free tier, monitored for the announced 2027 API-key requirement).

Legacy fallbacks — Azure Translator, WordsAPI (RapidAPI), and freeDictionaryAPI (Wiktionary data under CC BY-SA) are kept only as compatibility fallbacks behind the local dictionary and are disabled by default.

Moonshine Base (onnx-community/moonshine-base-ONNX, MIT) — primary in-browser speech recognition for speak-back practice.

Whisper base.en (onnx-community/whisper-base.en, Apache-2.0) — fallback in-browser speech recognition.

Translation models (Hy-MT2-1.8B, MADLAD-400-3B, NLLB-200-600M) run only on the administrator's local machine to prepare course translations. They are not part of the public service and never process end-user traffic. NLLB-200 is licensed CC-BY-NC-4.0 (non-commercial) and is never used to serve users.

Dictionary datasets are used strictly for in-app word lookup: definitions, IPA pronunciations, forms, and optional translations shown while you study.

We do not sell, redistribute, republish, or modify the underlying datasets, and we do not offer dataset downloads. Attribution for every source is maintained on this page.

If you believe an attribution is missing or incorrect, contact us and we will fix it promptly.