Open bilingual dictionaries for e-readers

Free bilingual dictionaries are poor because almost every generator uses a single source: Wiktionary. These are built by merging Wiktionary with the wordnets of the Open Multilingual Wordnet, which all share Princeton WordNet synset identifiers and can therefore be pivoted against one another. Quality metrics for every dictionary are published here — measured, not claimed.

How these are built

Every wordnet in the Open Multilingual Wordnet maps its words onto the same Princeton WordNet 3.0 synset identifiers. Once two languages are both mapped, a dictionary between them already exists implicitly:

wield  --(WordNet)-->  01095899-v  --(OpenWN-PT)-->  empunhar, manejar

So a pair with no direct bilingual resource anywhere — Spanish to Portuguese, say — comes out of the same pipeline as English to Portuguese. Wiktionary is then layered on top for the slang, idiom and inflected forms that wordnets lack.

Why coverage matters more than the entry count

Entry counts are easy to inflate and say little. What decides whether a dictionary is useful is whether tapping a word in the middle of a book returns anything — and that depends on inflected forms as much as on headwords. A dictionary with no inflection index fails on wielded however many entries it advertises. Every pair here is measured against a frequency list, counting inflections, and the result is on its page.

Catalogue generated 2026-09-11 · 53 dictionaries built.