Translate within extended enterprises. What does that mean? – Part I


AFNET, a non-profit society for boosting the digital transformation within vertical industries

AFNET is a French non-profit organization that promotes best practices for processes within the extended enterprises. This society introduces Internet in France in 1992 and was the owner for delivering Internet access to the very first users, mainly in universities.

Since this early age, AFNET has been continuing to promote good practices for enterprises within their ecosystem of suppliers, partners and key customers. These IT practices consist on standards, a common information system, and a partner framework for co-working in confidence.

Today, the best approach consists to leverage industry verticals to build open systems for digital transaction, content collaboration and product design.

Corporations are grouped within industry verticals and are highly dependent themselves. Just a few examples: Do you know that less than 20% of Dassault Rafale airplane parts are done by Dassault Aviation itself? Do you know that behind an Airbus air plane, there is a network of more than 1,200 different suppliers structured by more than 5 depth levels?

To be competitive and efficient within our globalized world, the European aerospace industry had organized itself with AFNET expert assistance to create a common information system named BoostAeroSpace: AirSupply as the platform for managing the supply chain, AirCollab as the platform for data communication exchange and Airdesign as the platform for product design. All these highly-secured platforms managed by the nonprofit organization and financed by all aerospace members make a strong asset that contributes to Airbus success.

Similar projects are in progress today for automotive, energy, travels, that leverage aerospace good practices. All of them share the common vision of the extended enterprise. Continue reading

Voyage en traduction automatique

« Mais enfin, papa, tu ne vas pas aller là ! La traduction automatique, ça marche pas ! c’est pourri …», ainsi s’exprime la fille de l’auteur, 19 ans. Trait générationnel, elle est spontanée, directe.
L’auteur se gratte le front, légèrement ébranlé. Dans la main, il tient sa proposition d’embauche chez le leader mondial de la « machine translation ».

« Mais pourquoi dis-tu ça ? » Continue reading

How does Neural Machine Translation work?

The representation of meaning in Neural, Rule-Based and Phrase-Based Machine Translation

In this issue of step-by-step articles, we explain how neural machine translation (NMT) works and compare it with existing technologies: rule-based engines (RBMT) and phrase-based engines (PBMT, the most popular being Statistical Machine Translation – SMT).

The results obtained from Neural Machine Translation are amazing, in particular, the neural network’s paraphrasing. It almost seems as if the neural network really “understands” the sentence to translate. In this first article, we are interested in “meaning,” that which gives an idea of the type of semantic knowledge the neural networks use to translate.

Let us start with a glimpse of how the 3 technologies work, the different steps of each translation process and the resources that each technology uses to translate. Then we will take a look at a few examples and compare what each technology must do to translate them correctly.

Continue reading