6/29/2023 0 Comments Editing machine transcriptionsThis means that the future is rosy for those translators who are happy to undertake post-editing machine translation. Indeed, promises of machine translation being mastered within three to five years have been circulating since the first successful automated translation was achieved in 1954.Īs such, poorly worded machine translations are likely to be around for many more years yet. While many of the tech giants are promising they are on the brink of cracking machine translation, nobody has yet managed it, despite some staggering advances in neural networks and deep learning. The future of post-editing machine translation Their presentation makes them tricky to translate using a machine and it can be more cost effective (and faster) to use a human to undertake the work and to typeset the resulting copy. The human element of this process is so extensive that it is likely to be slower to use a computer.įorms and other formatted documents are another challenge. Video translation then presents the additional challenge of needing to be time-stamped, in order that the translated copy marries up to the images on screen. Again, the content is not in the correct format for the computer to get to work on translating it and must be transcribed. Video translation and audio translation also do not lend themselves well to machine translation. As such, the business must pay for the transcription before the document can be translated and them post-edited – in which case it is likely cheaper and faster to simply ask a human translator to work from the original, hand-written version. The text on them needs to be transcribed into electronic format before it can be translated. Handwritten documents, for example, can pose serious problems for machine translation. Using a computer to undertake a translation relies on the original copy being in an acceptable format. Of course, few processes are without their flaws, and post-editing machine translation is no exception. Disadvantages of post-editing machine translation And, as the post-editing machine translation is usually faster than human translation alone, the price of the manpower element has the potential to reduce significantly. The machine element of the process can cost nothing, depending on the choice of software used. With a reduced need for human input, post-editing machine translation can also reduce the cost of translation services to businesses. This means that companies can obtain their translations more quickly than previously, thus enabling them to progress whatever projects are waiting on those translations. The combination of computer and human input can make for a faster process than human translation alone. Advantages of post-editing machine translationįrom a business perspective, post-editing machine translation has some distinct advantages. Full post-editing machine translation goes further, with the translator working to ensure that the copy matches the original stylistically as well as grammatically. The purpose is simply to ensure that the copy is comprehensible. Light post-editing machine translation involves correcting mistakes and reviewing language choices that the computer system has made. Translators agree in advance with their clients on which service is required. There are two types of post-editing machine translation: light post-editing and full post-editing. The human translator works on the translation produced by the machine in order to produce a final version that is linguistically correct and an accurate representation of the original document. Post-editing machine translation blends machine translation with human input. What is post-editing machine translation? The fact that the results of machine translation don’t compare favourably to human translation is giving rise to the next phase in the translation industry’s evolution: post-editing machine translation. With the promise of translating faster and more cost effectively than humans are capable of, it’s easy to see why machine translation appeals to businesses. Companies are trying out machine translation to see if it meets their needs. However, the reality is that nobody has yet managed to produce an app, algorithm or anything else that can translate as skilfully as a human.Įven so, the way that we translate is evolving. Over the past few years, it seems that every big player in the digital/tech sector has announced that they have mastered machine translation.
0 Comments
Leave a Reply. |