On ERT and MERT-Rings
نویسندگان
چکیده
منابع مشابه
Optimising Multiple Metrics with MERT
Optimisation in statistical machine translation is usually made toward the BLEU score, but this metric is questioned about its relevance to an human evaluation. Many other metrics exist but none of them are in perfect harmony with human evaluation. On the other hand, most evaluation campaigns use multiple metrics (BLEU, TER, METEOR, etc.). Statistical machine translation systems can be optimise...
متن کاملOn n-coherent rings, n-hereditary rings and n-regular rings
We observe some new characterizations of $n$-presented modules. Using the concepts of $(n,0)$-injectivity and $(n,0)$-flatness of modules, we also present some characterizations of right $n$-coherent rings, right $n$-hereditary rings, and right $n$-regular rings.
متن کاملOn generalizations of semiperfect and perfect rings
We call a ring $R$ right generalized semiperfect if every simple right $R$-module is an epimorphic image of a flat right $R$-module with small kernel, that is, every simple right $R$-module has a flat $B$-cover. We give some properties of such rings along with examples. We introduce flat strong covers as flat covers which are also flat $B$-covers and give characterizations of $A$-perfe...
متن کاملon semihypergroups and hypergroups
in this thesis, first the notion of weak mutual associativity (w.m.a.) and the necessary and sufficient condition for a $(l,gamma)$-associated hypersemigroup $(h, ast)$ derived from some family of $lesssim$-preordered semigroups to be a hypergroup, are given. second, by proving the fact that the concrete categories, semihypergroups and hypergroups have not free objects we will introduce t...
15 صفحه اولA Unified SMT Framework Combining MIRA and MERT
Translation sub-model is one of the most important components in statistical machine translation, but the conventional approach suffers from two major problems. Firstly, translation sub-model is not optimized with respect to any of automatic evaluation metrics of SMT (such as BLEU). The second problem is over-fitting to training data. This paper presents a new unified framework, by adding a sca...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: AL-Rafidain Journal of Computer Sciences and Mathematics
سال: 2004
ISSN: 2311-7990
DOI: 10.33899/csmj.2004.164095