Q: Before making your phrasebooks, I'd get actual speakers and translators to check your data, because your English / Spanish pair is nowhere near 80% accurate - more like 10% accurate or less! Most of the words claiming to be English are not English - they are proper Spanish words that some English speakers use in bilingual areas such as California and are not English words!! Try saying any of those in areas without many Spanish speakers such as Australia or England and no-one will understand you. (http://www.ezglot.com/common-words.php?l=spa&l2=eng&p=nKha)

A: English / Spanish pair was checked by a very competent Spanish translator and a fluent English speaker. So maybe that is why given set works better for Spanish speakers than English speakers. Phrasebooks will be made by actual speakers and translators of given pair of languages. We did not exclude proper names from our dataset. User validations of word pairs will give us much better estimate of the accuracy as perceived by common speakers (not professional translators and linguists). Also, note that English dictionary has hundreds of thousands of words, and we connect words that are similar in meaning and understandable, not only exact translations. One's perception of accurracy may depend on vocabulary.

Q: I'm writing a language learning book and I was wondering if I could use your words or word list (are single words copyrighted?) in my book. (http://www.ezglot.com/)

A: You can use word lists. Just please mention us in your book. Cheers!

Q: Northern Sami and Norwegian Bokmål are totally different languages. Norwegian belongs to the Indo-European languages and Sami belongs to Uralic languages aka Finno-Ugric languages. (http://www.ezglot.com/most-similar-languages.php?l=sme)

A: We analyzed orthographic and phonetic similarity between similar concepts, not exact translations or grammar. Our formula calculates a similarity to another language in relation to similarities to all other languages. Mandarin and Japanese may not be similar, but of the words they share with other languages analyzed, most of them they share among themselves! We used publicly available dictionaries which are neither complete nor 100% accurate. We did not exclude proper names from our dataset. When better and more complete data becomes available, our similarity formula will yield more accurate values. Our assumption was that no single language is isolated, and every language shares something with other language(s). In this regard, the number of shared words between two languages in proportion to all words they share with all languages, gives their similarity relative to similarities to other languages. Coverage formula is an indicator of how much public data was contributed to given language pair in relation to total coverage of dictionaires involving those two languages.

Word validations


Common words for English and French
look: not French, unless it's been recently borrowed.
must: not the preterite-present verb (the French for which is dois) but grape juice.
listen-liste: should list-liste. The French fol listen is écouter.
shit: the French is merde.
found: to found is fonder, foundry is related to fondre,
 but if what you're looking for is found, that's trouvé.
case-cause: case-cas and cause-cause.
drink: boire.
mine-miner,mina: mina is the past of miner.
person-persona: personne.
office: oficine.
system: système.
court: court-cour and short-court.
boss: don't know that in French.
sort: means luck or fate.
train-traîne: I think that's right if you're training someone, but the vehicle is train.

A: Thanks for feedback. Like is said on the site, accuracy of the data is only around 80%. Hopefully visitors will flag enough so it becomes more accurate :) We do not intend to currate ourselves millions of connections between 90 languages. For phrasebooks we will hire professionals for given language pairs, and we will offer access to verified word lists, under certain conditions.

Full data access

Q: How can I get a full data access?

A: You may get full access to data in specified language pair of your choice if you provide your contact info, a website if you have one, languages you speak or learn, and other info you think may be relevant. Depending on your qualifications you may be offered one of several modes of accessing full data.

Full data access - bad requests

Q: I would appreciate having access to the full listing of your German/English common words, as well as the full listing of your Spanish/English common words. (http://www.ezglot.com/common-words.php?l=deu&l2=eng)

A: Why? We would appreciate knowing more about you.

Full data access - better requests

Q: I'd really like access to the databases to see where Greek and Latin overlap please, :) I think this could be bonus vocab as I learn more Latin. (http://www.ezglot.com/common-words.php?l=ell&l2=lat)

A: Indeed, it would make learning a bit easier! You need to tell us more about yourself though. We invested a lot of time and effort into this project and don't want to give away data just like that.

Q: I'm a hyperpolyglot and I'm interested in testing out this page, to see if it actually learns me a language faster than i usually do. I need the full access to this page before continuing. (http://www.ezglot.com/common-words.php?l=ara&l2=spa)

A: See above.

Full data access - good requests

Q: I'd like to inquire about gaining full access--is this something you offer after placing a donation, or through subscriptions? I might also be interested in contributing, depending on what qualifications you require. My native language in English, and I'm intermediate in Polish. (http://www.ezglot.com/common-words.php?l=eng&l2=pol)

A: For contributions we prefer bilinguals... Just curious, how much is ~1000 common words for English/Polish worth to you?