Industry Observation| Wang Huashu: Development of Translation Technology in the Big Data Era and Its

Industry Observation| Wang Huashu: Development of Translation Technology in the Big Data Era and Its Enlightenment

Source: YEEWORLD.COM       Author: Wang Huashu        Date: 2016/08/30

In the age of big data, data is everywhere. The soaring data will give rise to the upsurging demands for communication and further the exploding demands for language services. Big data is a comprehensive technology, reflecting the technical nature of the society. An important aspect of the technical society is that technical factors play an active role in it and among these factors technical development and innovation are dominant which will have an immeasurable influence on the development of the translation industry. In the language services sector, information that could hardly be quantified in the past will be stored and processed in the form of data, and a large amount of complex projects to be translated gradually surface. Therefore, digging for the value hidden in data to create the blue ocean in the language services sector is an irresistible trend in the translation industry. Traditional translation researchers are confined to the study of language and texts, without realizing the important role that translation technology is playing in today’s business environment. In the meantime, it is almost impossible to use traditional translation theories to describe and explain new translation technology phenomena and  activities at the present day. Whether we are prepared or not, the rapid development of translation technology in big data era will bring about significant changes to translation studies and translation education on a global scale.

1.      Language Services Revolution in the Big Data Era

The language services industry is generally perceived to cover four business areas, i.e. translation and localization services, language technology tools development, language education and training as well as multilingual information consultation. This industry cannot develop without the high-speed processing of mass information. Against the backdrop of economic globalization, with the exponential growth of information, the amount of data generated in the past two years is approximately equal to the total amount of all previous times in human history. Knowledge growth and differentiation have far exceeded the maximum capacity of human beings. Therefore, in the information era, the society cannot develop without information processing technology. Under this circumstance, big data computing technology comes into existence and offers solutions to the technical problem that traditional computing methods cannot complete the analysis and processing of large data in reasonable period of time. Big data technology and natural language processing technology based on statistical approaches have made major breakthroughs in speech recognition, machine translation, semantic retrieval, etc. (Tang Zhifang, Yu Yang, 2015) Over the past few years, the market of language services and technology has been developing, with the annual growth rising year by year, from USD 25 billion in 2009 to USD 40.27 billion in 2016. This revolution brings the language services industry into a brand new information era: language services are no longer limited to interpreting and translation, but become increasingly diversified over time. Big data brings many new types of businesses and significant changes to the structure of language services market. In 2016, the most important businesses in the industry are still traditional translation and on-site interpreting with their total market share rising to 73% from 57% in 2013. However, compared with 2013, 2016 witnessed the emergence of new businesses including conference interpreting (3.32%), mobile apps localization (0.51%), game localization (0.54%), search engine optimization (0.35%), and subtitle translation (1.08%). Despite their small market shares, these emerging sectors are relatively more stable (CSA, 2016). In the meantime, the modes of language services also vary to the diversified market. For instance, as real-time multilingual communication and translation are often required in cross-border e-commerce, a real-time, dynamic and fragmented micro language service mode appears on the scene. In this case, diversified language strategies should also keep adapting to the market.

2.      The Development of Translation Technology in the Big Data Era

In the age of big data, the world is consisted of data and expressed by data. The data that human beings have known is just a tip of the iceberg, while a large amount of data are still not fully tapped, understood and applied. Facing this massive, undeveloped and unstructured data, to dig more data that is meaningful and valuable to certain industries, we are in urgent need of modern language processing technology. Driven by new technologies, the emerging language services market features massive, diversified, fragmented, multimodal and instantaneous data. These features make language technologies a crucial support for the market. In the age of big data, taking translation as an example, translated texts as products may be labeled as data and many factors may be tracked, such as the emergence of original texts, the continuous vitality of translated texts, and translators’ style and the version management, cross-national dissemination, audiences and dissemination effects of translated texts. All these data can form a big translation database, which will have a profound influence on translation education and research.

With the development of information technology and in particular the driving force of cloud computing and big data technology in the past few years, speech recognition, translation technology and translation platform technology have had a great leap forward. In the big data era, corpus resources are getting richer and speech recognition enjoys rapid development. Iflytek Co. Ltd has developed many intelligent language technologies, such as voice dictation, voice input, voice translation, voice learning, conference dictation, public opinion monitoring. Many developers of translation tools represented by SDL have also developed many Internet-based tools like technical writing, translation memory, term management, voice recognition, automatic quality assurance, translation management, which are used widely in translation practices in the industry. In addition, computer Aided Translation (CAT) software has also made a huge progress, growing from single versions to versions with online collaboration and cloud functions, from single PC platforms to diversified smart ends. Ecologically integrated crowd-sourcing translation platforms including Flitto, TryCan and Onesky have also gained benefits from big data technology and are thriving. Take Trycan, developed by Zhongye Science and Technology Co., Ltd., as an example. Relying on the big data from the Internet, it combines language environment and geological and other factors in different countries. In addition, with tens of thousands of online part-time translators, a multi-layer senior translator checking systems, and especially tight time limit, it makes sure that all types of translations can be finished in one minute. This software has changed the modes of machine translation and human translation and made translation more human.

3.      Translation Education in the Big Data Era

Compared with translation technology education, traditional translation education started earlier in China. In 2006, the Ministry of Education authorized three universities to set up Bachelor of Translation and Interpreting (BTI); in 2007, 15 universities were approved to establish Master of Translation and Interpreting (MTI); in 2011, 158 universities had MTI and 42 had BTI; in 2012, 159 had MTI and 106 had BTI; in 2016, 206 had MTI and 230 had BTI. On such an occasion, Translation professionals should have been able to satisfy the market demands. However, the rapid developing technologies ask for translators with comprehensive, multi-dimensional and diversified abilities, so graduates provided by colleges and universities can hardly meet enterprises’ recruitment requirements. Under this circumstance, MTI in Beijing University started to launch curriculums like Translation Technology and Localization Construction in 2009, and pioneered in  Language Services Management in 2013. By far, it has provided a few groups of qualified professionals for enterprises. Yet the limited  number of graduates is still insufficient to meet the growing demands from enterprises. From 2014 to 2015, Beijing Language and Culture University, Guangdong University of Foreign Studies and  Xi’an International Studies University set up the major of Localization, aiming at training professionals in need in the localization services market. Besides, according to local regional and economic development, a few universities have carried out customized professional training strategies in collaboration with international enterprises.

In the big data era, machine translation, CAT, smart voice transcription and recognition are developing rapidly. Translation technology has been transferred from the desktop to the cloud and to everywhere. As the language services market is in constant changes, its requirements for professionals are also different from those in the past. Compared with the traditional translation industry, modern translation differs significantly in translation objects, forms, procedures, methods and environment. These changes should also be fully considered in translation education. Only in this way can we have professional who can meet the demands of the market. In the ecological environment of translation, the technical processing of translation involves a series of procedures, including translation, proofreading, QA, management and typesetting technologies, and corresponding personnels from professional translators, proofreaders, QAs, project managers and typesetters. In this systematic model, the technological demands in the translation system are in line with the training objectives of the professional training system, which requires the close connection between translation education and market development. Higher education institutions should first analyze the market and set reasonable professional training goals. Then they should adjust the content of translation education, adding translation technology urgently required by the market to the curriculums, in order to nurture language services professionals who know both translation and translation technology. For instance, with the assistance of big data, machine translation technology has made major progress. The machine translation of Microsoft, by using the deep neural network technology, can achieve an accuracy rate between 80% and 90% in certain areas. After the preliminary translation by the machine translation system, human editing and proofreading are all we need to finish the translation task, which is referred to as the machine translation post-editing mode. Accordingly, translators need to know how to “cooperate” with the machine and how to do post-editing efficiently. This part of knowledge is essential in translation education, because this is how we can make sure the professionals we train satisfy enterprises’ demands.

In addition, translation technology education is different from traditional translation education. In the former, modern education technology and platforms (such as Moodle course management system, Virtualclass system and MOOC education) are used to incorporate latest outcomes of modern education into translation technology curriculums, so as to promote innovation in translation education and keep it advance with the times.

4.      Summary

In the big data era, translation technology is essentially the integration of translation humanity and technology under digital humanism. The two factors influence, interact and integrate with each other. Translation technology constructs and facilitates translation activities and society. In the new era, translation practitioners can find no way to escape from translation technology. We should keep open to the developing new technologies and fully aware the relationship between the humanity and technicality of technologies. We should make full use of the advantages of modern translation technology, keep in pace with the developing market and adjust professional training strategies and syllabus, so as to make modern language services talents with comprehensive qualities who can keep up with the changing time and meet the demands of the market.

(The author is a teacher at Guangdong Collaborative Innovation Center for Language Research and Service in Guangdong University of Foreign Studies. This article was first published in East Journal of Translation and has been authorized here by the author)