49.7% Better Than Google NMT For English-to-Italian IT Software

This English-to-Italian IT Software translator uses a Slate Desktop personalized SMT that’s 49.7% 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.

sourceen
targetit
subject domainit-software
estimated corpus size50,000
segments per representative set2,292
words per source segment15
words per target segment16

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.

 Google
NMT
Slate Desktop
SMT
words per MT segment1616
BLEU score41.1981.35
exact MT match (count)1071193
exact MT match (percent)4.7%52.1%
words per exact MT match (count)611
filtered BLEU score (no exact MT matches)40.2470.99
segments requiring edit (count)2,1851,099
character edits per segment2822
total character edits61,18024,178

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.

enit (Google and translator)
The results contain reflective objects and have been rendered with reflection depths of 2, 4 and 8I risultati contengono oggetti riflettenti e sono stati renderizzati con profondità di riflessione di 2, 4 e 8
The only editable data here are the six fields at the top of the dialog box.Gli unici dati modificabili qui sono i sei campi nella parte superiore della finestra di dialogo.
If needed, you can redefine the direction of this cursor projection.Si consiglia di duplicare il traduttore predefinito prima di modificarlo.

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.

enit (Slate and translator)
In Edit mode, each member (Scheme, Frame, Panel, Accessory, Junction) has its own Dialog box and its own Tool, enabling you to change settings and place new Members in this Curtain Wall.In modalità Edita, ciascun elemento (Schema, Profilo, Pannello, Accessorio, Giunto) ha la sua propria finestra di dialogo e il suo Strumento, cosa che permette di cambiare le impostazioni e posizionare nuovi Elementi in questo Curtain Wall.
In case of a slanted wall, however, you might prefer a more realistic display of the cut door/window: choose either Projected or Projected with Overhead to see all parts of the window in a slanted wall.Nel caso dei muri inclinati invece si può preferire una visualizzazione più realistica delle porte/finestre sezionate: scegliere Proiettato o Proiettato con parte superiore per vedere tutte le parti delle finestre nei muri inclinati.
Note that if you drag a Master Layout onto a Subset in Layout Settings, the automatic reservation (if available) will reserve every sub-element in that Subset (all Layouts and all Drawings).Notare che se si trascina un Layout Master su un Sottoinsieme in Impostazioni Layout, la riserva automatica (se disponibile) riserverà tutti i sottoelementi in tale Sottoinsieme (tutti i layout e tutti i disegni).
Advertisements