This English-to-Italian regulatory translator uses a Slate Desktop personalized SMT that’s 30.8% more productive than Google NMT.
A Slate Rocks customer (a translator) created a translation engine with his or her translation memories (TMs) using a personal computer. This page describes the engine and compares the translator’s Slate Desktop experience to an experience using Google’s new-and-improved neural machine translation (NMT) technology. You can read the entire report with thirth (30) more customer experiences by downloading the full report Study of Machine Translated Segment Pairs.
Slate Desktop Engine Details
The customer started with translation memories in the language pair and industry of his or her work, totaling the estimated corpus size number of segments. Slate Desktop cleaned the TMs, prepared a training corpus and built the engine. Note that these processes typically runs overnight. During that processing, Slate Desktop also extracted a representative set consisting of randomly selected segments from the training corpus.
|estimated corpus size||979,150|
|segments per representative set||3,381|
|words per source segment||27|
|words per target segment||30|
Benchmark Score Comparison
The segment pairs in the representative set are representative of the translator’s daily work. By focusing on one translator’s experience, these scores indicate a level of work reduction this customer will likely experience in his or her daily work using the respective MT system (Google or Slate Desktop) with 95% confidence.
|words per MT segment||28||28|
|exact MT match (count)||131||1132|
|exact MT match (percent)||0.039||0.335|
|words per exact MT match (count)||12||20|
|filtered BLEU score (no exact MT matches)||34.91||66.03|
|segments requiring edit (count)||3,250||2,249|
|character edits per segment||63||45|
|total character edits||204,750||101,205|
These scores indicate this customer using Slate Desktop will likely spend significantly less time editing MT suggestions than if he or she were using Google for this work. This is because Slate Desktop creates engines with the customers translation memories and optimizes them to predict how the customer translates. While on the other hand, Google optimizes its NMT service for millions of customers with countless demands.
Google’s three (3) longest exact MT matches
The exact MT match (count) in the Benchmark Scores table (above) is the number of segments that Google NMT successfully matched to the translator’s actual work, i.e. Google successfully predicted the translator’s actions. The three segments in this table are the exact MT match segments with the longest length. This translator can expect to experience these kinds Google NMT results while translating these kinds of project.
|en||it (Google and translator)|
|The Sub-Fund will close upon closure of the last active Share Class.||Il Comparto chiuderà alla chiusura dell’ultima Classe di Azioni attiva.|
|The Hang Seng index represents the reference indicator of the Sub-Fund.||L’indice Hang Seng rappresenta l’indicatore di riferimento del Comparto.|
|Le attività sono state svolte dalla società consortile ABC 991.||The activities were carried out by the consortium company ABC 991.|
Slate’s three (3) longest exact MT matches
The exact MT match (count) in the Benchmark Scores table (above) is the number of segments that this translator’s Slate Desktop engine successfully matched to the translator’s actual work, i.e. Slate Desktop successfully predicted the translator’s actions. The three segments in this table are the exact MT match segments with the longest length. This translator can expect to experience these kinds Slate Desktop results while translating these kinds of project.
|en||it (Slate and translator)|
|By mentioning rules of transparency and non-discrimination, the budget law for 2004 has codified a principle that was already underlying the original version of Article 1, section 2, and intended to allay the risk that the sale take place at a price that is below market price, favoring the purchaser from a competitive point of view, in violation of the law on the matter of State aid.||Con il richiamo a regole di trasparenza e non discriminazione, la legge finanziaria 2004 ha codificato un principio già sotteso alla versione originaria dell’art. 1, comma 2, ed inteso a scongiurare il rischio che la cessione abbia luogo ad un prezzo inferiore a quello di mercato, favorendo l’acquirente dal punto di vista concorrenziale in violazione della normativa in materia di aiuti di Stato.|
|If the Subfund is not going to be liquidated because of such a termination at short notice of the overall derivative structure, but rather the current counterparty to such an overall derivative structure is to be replaced by a new counterparty, then during the transition period it may also occur that investors do not participate.”||Qualora il Comparto non venisse liquidato a seguito della suddetta risoluzione della struttura complessiva dei derivati con breve preavviso, bensì la controparte attuale in detta struttura complessiva dei derivati fosse sostituita da una nuova controparte, nel periodo di transizione potrebbe accadere che gli investitori non partecipino.”|
|Based on an inspection of the amounts of the service components identified by you in Annex 1 to your notice referred to above, ABC is currently using 14 housing sites (hereinafter the Sites) and 25 Ethernet interface connections with a total capacity of around 2.5Gb/s (hereinafter, the Connections).||In base alla verifica delle consistenze delle componenti di servizio da Voi individuate nell’Allegato 1 alla Vostra comunicazione di cui in oggetto, risultano in uso ad ABC alla data di oggi 14 siti in housing (di seguito, i Siti) e 25 collegamenti con interfaccia Ethernet, di capacità complessiva di circa 2,5Gb/s (di seguito, i Collegamenti).|