Sources and corrections
Everything the Spanish deck is built from, and how to tell us when something is wrong.
Spanish deck (version 1.0.0)
| Part | Source | License |
|---|---|---|
| Frequency ranking | wordfreq 3.1.1 by Robyn Speer (Spanish data: Wikipedia, OPUS OpenSubtitles 2018, SUBTLEX, NewsCrawl 2014, GlobalVoices, Google Books Ngrams 2012, OSCAR web text, Twitter, Reddit) | Data: CC BY-SA 4.0. Code: Apache 2.0. |
| Grouping forms under headwords | simplemma 2.0 lemmatizer, plus hand corrections | MIT |
| Near-synonym check for quizzes | Princeton WordNet 3.0 via NLTK | WordNet License (permissive) |
| Meanings, notes, example sentences, translations | Written for this project with AI assistance and checked automatically. Not yet reviewed by a native speaker. | Part of this site |
Because the ranking is derived from wordfreq data, the ranked headword list is shared under CC BY-SA 4.0 with attribution to wordfreq. No proprietary flashcard decks or restricted dictionaries were copied.
Software
- Scheduling: ts-fsrs (MIT), implementing FSRS.
- Accounts: Better Auth (MIT).
- Site: Next.js (MIT). Fonts: Newsreader and Public Sans (SIL Open Font License).
Known limitations
- Frequency is measured on written and subtitle text up to about 2021, not on everyday conversation.
- Each entry shows its most common senses, not every sense.
- Example sentences are short and simple, written to show the word in use. They aren’t taken from real texts.
Report a correction
Found a wrong meaning, an unnatural example, a spelling error, or a regional note that’s off? Send it here. Reports go into our review queue.