Some unexpected pitfalls are lurking in the process of disambiguation of some expressions or locutions. In the above example, the French locution ‘en train de’ can be translated :
- either into in trenu da (by train from)
- or into in traccia di (in the process of)
Now updating the efficiency open test’s protocol, targeted at measuring the accuracy of the French to Corsican translation. It is based on the wikipedia first 100 words of the ‘article of the day’, which changes everyday. The scoring will result from the 10 last tests performed. The test will be performed every saturday. Here is the first one. Scoring = 1 – (6/138) = 95,65%. One vocabulary error: the verb ‘égaler’ (to equal). It should also read ‘da u 1923 à u 1939‘, ‘u 16u‘ instead of ‘u 16e’, ‘a prova maestra‘.
The test is termed ‘open’ in the sense that:
- its protocol is clearly described
- the ‘article of the day’ of the french wikipedia is pseudo-random (there is room for discussion here ; arguably, it does not consist here of perfect randomness, but it can be considered that it provides an acceptable level of randomness)
- it is ‘open’ since it can be verified via the ochakko translator for android app
Translating this series of acid names is not as easy as it could seem at first glance. In effect, each acid name is composed of three consecutive ambiguous names:
- ‘l’ is ambiguous between the masculine (u, the) or feminine (a, the) definite article
- ‘acide’ is ambiguous betwwen acidu (acid, masculine singular noun), acitu (acid, masculine singular adjective) or acita (acid, feminine singular noun)
- ‘daturique’, etc. are all ambiguous since that can be either masculine singular (daturicu, daturic) or feminine singular (daturica, daturic) adjectives.