Five years of progress in machine translation

People who write on the web have an interest in machine translation. If it worked communication could suddenly become much more international. Five years ago yesterday I wrote a post about the state of the art.

Of course, machine translation is still a developing technology, somewhere about the level of voice recognition 10 years ago is my estimate. As a test of the new service I supplied a paragraph chosen pretty much at random from the Amos commentary, the first from a page about city gates:

In the Ancient Near East city gates were neither merely entrances, nor only used for military protection. As a potentially weak point in the defenses, the gates of Israelite walled cities typically had three chambers giving four sets of “doors” and defended spaces between. The gates of Hazor (left – plan above) and Gezer from the time of Solomon show this triple construction.

This gave the almost comprehensible French:

Dans la ville antique du proche Orient les portes n’étaient ni simplement des entrées, ni seulement utilisé pour la protection militaire. Comme point potentiellement faible dans les défenses, les portes d’Israelite ont muré des villes ont typiquement eu trois chambres donner quatre ensembles d'”portes” et d’espaces défendus entre. Les portes de Hazor (laissé – projetez ci-dessus) et de Gezer dès l’exposition de Solomon cette construction triple.

The Iron Age gate at Hazor

The Iron Age gate at Hazor

Either Google translate is more than five years behind that crowd, or the art has gone backwards in the last half decade, here’s Google’s 2010 rendering of the same text:

Dans l’ancienne ville près des portes-Orient ne sont ni seulement les entrées, ni seulement utilisés pour la protection militaire. Comme un point de faiblesse potentielle dans la défense, aux portes des villes d’Israël paroi généralement avait trois chambres donnant quatre séries de “portes” et défendu les espaces entre les deux. Les portes de Hatsor (à gauche – plan ci-dessus) et Gezer du temps de Salomon montrer cette construction triple.

Hmm… well let’s follow the same methodology, now from “French” into German:

In der alten Stadt in der Nähe des East Türen sind nicht nur Eingänge oder nur für den militärischen Schutz. Als Ausgangspunkt der mögliche Schwäche in der Verteidigung vor den Toren der ummauerten Städte Israels in der Regel hatte drei Zimmer gibt vier Gruppen von “Türen” und verteidigte die Zwischenräume. Die Tore von Hazor (links – über Plan) und Gezer Zeit Salomos zeigen, dass der Bau verdreifachen.

Which of course means (in “English”):

In the old city near the East doors are not only inputs or only military for protection. As a point of potential weakness in the defense at the gates of the walled cities of Israel usually had three rooms are four categories of “doors” and defended the gaps. The gates of Hazor (left – show on plan) and Gezer time of Solomon, that tripled the building.

Sadly we still have some way to go,  though, the back translation has actually improved! This was the original version:

In for the military protection antiken of the Near East the doors were neither simple from the entrances nor only use city. When potentially weak point in the defenses the doors have typically three chambers immured of Israelite of cities had to give four whole of “doors” and from areas to which were defended, between. The doors of Hazor (calmly – above you plan) and of Gezer of the exhibition of Solomon to this three-way construction.

Maybe in another five years we’ll have a usable machine translator at our fingertips, and the World Wide Web will really be starting to become world-wide, in the meantime the majority of the world’s population will continue to learn English, providing jobs for Kiwis on their great OE  ;)

3 comments on “Five years of progress in machine translation

  1. mgvh

    Hi, Tim. Results doubtless depend on the translation engine. I ran your text through Google Wave using the “translatey” app, and here is what I got in the English > French > German > English. My French is worthless so I can’t evaluate that, but the German appears comprehensible, and the final English isn’t terribly far off.

    Dans l’ancien Proche-Orient portes de la ville ne sont ni simplement les entrées, ni seulement utilisé pour la protection militaire. Comme point de faiblesse potentielle dans les défenses, les portes des villes d’Israël avaient muré généralement trois chambres donnant quatre séries de “portes” et défendu les espaces entre les deux. Les portes de Hatsor (gauche – plan ci-dessus) et Gezer du temps de Salomon montrer cette triple construction.

    Im alten Orient Toren der Stadt sind nicht nur Eingänge, oder nur für den militärischen Schutz verwendet werden. Als Ausgangspunkt der mögliche Schwäche in der Verteidigung wurden die Tore der Stadt von Israel in der Regel drei Zimmer Mauern gibt vier Gruppen von "Türen" und verteidigte die Zwischenräume. Die Tore von Hazor (links – über Plan) und Geser in der Zeit Salomos zeigen, dass Triple Bau.

    In the ancient Near East gates of the city are not only inputs, or only be used for military protection. As a starting point of potential weakness in the defense of the city gates of Israel were usually three rooms walls are four categories of "doors" and defended the gaps. The gates of Hazor (left – above target) and Gezer in the time of Solomon to show that triple construction.

  2. Pingback: Gentle Wisdom» Blog Archive » Google Animal Translate