Machine translation, that was one of the most daring dream of artificial intelligence in the seventeenth century, came true in the late twentieth century. Although these translations, especially literary translations, were not as good as those done by human translators, they were proper for technical purposes such as medical reports.
The term machine translation refers to computer systems that perform translations with or without human assistance. Machine translation is different from Computer-Aided Translation (CAT) tools that help translators access, for example, online dictionaries and vocabulary databases.
Machine Translation vs. Human Translation
The boundary between machine-assisted translation and human-assisted translation is not yet well defined, and the term CAT tools can be used to refer to both, but the fully automatic translation process plays a key role in machine translation.
The use of machine dictionaries to overcome language barriers was first suggested in the seventeenth century. George Artsrouni and Petr Smirnov-Troyanskii worked individually on ways to create MT in the 1930s, with Smirnov-Troyanskii laying the groundwork for what was needed in an MT system. Namely, Smirnov-Troyanskii suggested the system would require an editor familiar with a source language to convert words to base forms before sending them to a machine to turn them into equivalent forms in the target language.
A few years after their invention, the possibility of using computers for translation was first discussed by Warren Weaver and Andrew Booth. And a few years later, machine translation research began at the University of Washington, the University of California, and the Massachusetts Institute of Technology. In 1951, the first machine translation research conference was held.
In the 1950s, machine translation became a reality in research. In the mid-1960s, research groups were formed in many countries around the world, including many European countries, China, Mexico, and Japan. Many of these groups were unable to achieve their breakthrough and disbanded without any success. But some of them, including the research group at the University of Grenoble in the 1960s achieved amazing results.
In 1964, federal machine translation sponsors in the United States set up the Automated Language Processing Advisory Committee (ALPAC) to focus more on prospects. The committee’s seminal report in 1966 showed that machine translation was less accurate and twice as expensive as human translation, and that there was no promising prospect. As a result, the committee stopped investing in machine translation research and instead, recommended the development of machine tools for translators. The project was the culmination of more than a decade of machine translation research in the United States, and machine translation was abandoned for years as a failed project.
SISTRAN and Logos
Finally, a new machine translation system appeared in the 1980s, and research expanded. This was exactly a decade after the ALPAC project. SISTRAN and later Logos were the most successful of these systems.
In the second half of the 1980s, interlinguistic systems came to the fore again. The dominant framework for machine translation research until the late 1980s was based on various linguistic rules, such as syntactic rules, Lexical rules, and morphological rules. However, from the 1980s onwards, corpus methods as an important innovation in the field of machine translation were introduced.
Although the goal of machine translation may be to provide high quality translations, the output is often revised, which is called “post-editing”. Of course, the human translations may also be edited by another translator before publication. Nevertheless, computer-generated errors are different from human translator errors. Sometimes the machine translation is not edited at all or slightly edited. The quality of machine translations may be improved with the development of more advanced methods or the imposition of certain restrictions on output.
درباره مریم پورگلوی
مریم پورگلوی، مترجم، مولف، مدرس دانشگاه و محقق اهل ایران است. او فارغ التحصیل مقطع کارشناسی ارشد مطالعات ترجمه است و به ویژه به مباحث مرتبط با ترجمه و فناوری و تاثیر فناوری های نوین بر روی صنعت ترجمه می پردازد.نوشتههای بیشتر از مریم پورگلوی