Machine translation is definitely a high-frequency word in the tech media in 2018. I have seen the news that Chinese players have won good places in international competitions such as WMT and IWLST, and I have also seen horror titles such as "Uncovering the Secrets Behind the Simultaneous Interpretation of the China International Import Expo: Not Humans, but Robots"; I have seen real-time subtitle projection machines with the same The rumors are amazing, and I have also experienced the "human-machine coupling" screen refresh;
I have seen the press conference news business email list of the translation machine translation egg translation headset, and I have also waited for the emergence of a hundred-yuan price butcher... However, as a curious and persistent follower of translation products, Watcher is more concerned about how machine translation, a technology that has become more and more mature in the past two years, has evolved at the product level under this buzz. Which scenes did it penetrate? What is the trend? Let's take a picture first: Machine translation + scene is the most interesting part. Observer divides them into three categories: information release, information acquisition, and information exchange.
Of course, this classification is not absolute. For example, video translation and document translation can be attributed to both information release and information acquisition, and there is a trend of integration. The following is a detailed description of this picture: 1. Open translation system The most common form is this online translation: For this type of product, in addition to the silent increase in the number of languages, various systems are secretly competing, and there are two major changes: One is the translation of documents that appeared almost overnight, and even the old-fashioned Google Translate uncle has been revised; Second, companies such as GTCOM, Xinyi, Atman, etc