Research On Collaborative Translation Tools Based On Cloud Translation Platform

With the increasingly popular of global communication and the rapid development of information science and technology, the language communication obstacle problem between different race and different state is increasingly obvious, as the communication between different languages become more and more important, the translation industry also got developed quickly. The current translation market is still human-based translation, although human translations can fully be up to the standard of translator, but it is inefficiency and high cost, so computer-aided translations is commonly used. Cloud translation platform is a collaborative translation platform based on the cloud platform, applies advanced computer and language processing technology to translation sector, and can realize the collaborative translation functions, organizing multiple spatially distributed users to complete one translation task, in order to improve the efficiency of the whole translation industry.This paper mainly studied how to inject the project team management into the collaborative translation tools based on the cloud platform, so as to better coordinate the relationship between different translators and translation project, and the aided translation input method will be introduced to collaborative translation platform, as a bridge between the translator and the collaborative translation platform, to effectively improve the efficiency of different translators. At the same time we realize the terms detection and recognition system and translation memory system, to not only improve the efficiency of different translators, but also can solve the problem that translation content repetition and lack of communication between different translators. In this paper, the main work and innovation involved as following:(1) For terms detection and recognition system, we treat terms identification problem as a classification problem, and introduce the method of ensemble learning at the same time, give full consideration to the situation that professional terms and general phrases distribution imbalance, through integrate the classification results of the multiple classifier, to minimize the classification error, significantly improve the accuracy of the terminology recognition, and can improving the consistency among the translation group.(2) We introduce an aided translation input method into the project team management function of the collaborative translation tools based on the cloud platform, to better coordinate the relationship between the different translators and translation projects. At the same time, to make translators work less and create more, saving time and improving the efficiency of translation.