Machine Translation (MT) has come a long way in recent years, but it still has its limitations. While MT can save time and provide a rough draft of a text quickly, it is not always accurate and often requires post-editing to improve its quality.

This blog aims to shed light on the essential process of post-editing in machine translation, emphasizing its critical role in enhancing the quality and reliability of translated content. By exploring the advantages, challenges, and best practices associated with post-editing, readers will gain a comprehensive understanding of how it contributes to effective global communication.

Unlocking the Potential of Machine Translation with Post-Editing

Definition and Scope of Post-Editing in the MT Process

Post-editing is the process of refining machine translation (MT) outputs by human editors to ensure the translated text is accurate, culturally relevant, and contextually appropriate. This involves correcting errors, enhancing readability, and adjusting the tone to suit the target audience. The scope of post-editing can vary from light edits, focusing on major errors, to full post-editing, which aims for a high-quality, publishable output nearly indistinguishable from human translation.

The Synergy between AI Capabilities and Human Expertise

The combination of AI-driven machine translation and human expertise creates a powerful synergy that leverages the strengths of both. While MT provides speed and efficiency, capable of translating vast amounts of text in seconds, it often lacks the ability to fully grasp cultural nuances and contextual subtleties. Human post-editors bridge this gap, applying their understanding of cultural contexts, idiomatic expressions, and specialized terminology. This collaboration ensures translations are not only quick and cost-effective but also accurate and culturally sensitive.

Pros and Cons of Post-Editing Machine Translations

Benefits: Time Efficiency, Cost-Effectiveness, and Scalability

Post-editing machine translations offer significant benefits, including time efficiency, as they shorten the translation process by providing a preliminary text that only needs refining. They’re also cost-effective, reducing the need for full translations from scratch by human translators, and scalable, allowing businesses and organizations to handle larger volumes of content that would be cost-prohibitive with human translation alone.

Challenges and Pitfalls in Post-Editing:

Need for Skilled Editors, Maintaining Consistency, and Understanding Context

However, this process faces challenges such as the need for skilled editors who are not only proficient in the language but also adept at understanding nuances and context. Maintaining consistency in terminology and style across large volumes of content can be difficult. Additionally, fully grasping the context and cultural subtleties in the original text requires a deep understanding that goes beyond linguistic knowledge.

Best Practices for Effective Post-Editing

Guidelines for Post-Editors to Enhance Translation Quality

To ensure high-quality translations, post-editors should follow guidelines such as understanding the purpose and target audience of the text, maintaining the author’s voice and style, and using translation memory and glossaries for consistency. They should also be aware of the cultural nuances and preferences of the target audience, adapting the translation to fit cultural contexts.

Tools and Technologies that Support the Post-Editing Process

Several tools and technologies support the post-editing process, including translation management systems (TMS), which streamline the workflow, and computer-assisted translation (CAT) tools, which provide access to translation memories and glossaries. AI-powered quality assurance tools can also help identify potential errors and inconsistencies, making the post-editor’s job more manageable.

Leveraging Professional Translation Agencies for Superior Post-Editing

While embracing the synergy between AI-driven machine translation and human expertise offers numerous benefits, engaging professional translation agencies elevates this process to new heights. These agencies play a pivotal role in global communication, providing not just post-editing services but a comprehensive suite of translation solutions that address the complexities of language with unmatched precision and cultural sensitivity.

Conclusion

The integration of post-editing in the machine translation workflow is not just an optional step; it is a necessity for overcoming the limitations of current AI technologies in understanding the depth of human languages. By combining the speed and efficiency of machine translation with the nuanced understanding of human editors, we can unlock the full potential of global communication.

Professional translation agencies play a key role in this process. They use the latest technology and the expertise of skilled editors to ensure translations are not only fast but also of high quality and culturally appropriate. As our world becomes more connected, the collaboration between technology and human expertise in translation is becoming increasingly important. This partnership is crucial for overcoming language barriers and helping us understand each other better, no matter where we are in the world.

The post Enhancing Global Communication: The Critical Role of Post-Editing in Machine Translation appeared first on Globalization Partners International.

What is Machine Translation?

Machine Translation is the process of translating content from one language to another, without the intervention of any human being.

Throughout history, there has always been a need for automatic translation without human intervention. The first experiments in machine translation date back to the late 1950s, when IBM, in collaboration with Georgetown University, translated more than 60 words from Russian to English.

About 60 Russian phrases related to political, legal, mathematical, or scientific topics were entered into the machine, which automatically translated them into English.

It wasn’t until the early 2000’s that the necessary hardware and software for more consistent translation became available.

Why is MT hard?

Israeli mathematician and machine translation pioneer Bar-Hillel presented a problem of how a translation system would deal with the phrase “The Box is in the Pen.”

The problem here is clear: The word “pen” has more than one meaning. It can mean “pen,” a writing tool, and at the same time, it can mean “playpen” for children.

To make a correct translation from one language to another, the system must determine which of the two uses of pen is the most appropriate.

«Little John was looking for his toy box. Finally, he found it. The box was in the pen. John was very happy»

A human reader will understand that the word “pen” refers to the playpen and not to a pen for writing.
The first sentence indicates “box” as a “box” that contains the toys. The reader is already aware that the box is much larger than a pen, so the first interpretation is automatically excluded without the reader having to think about it.

Where do we stand today?

The development of the Internet, together with globalization, produced a great demand for translation services and machine translation.

Global businesses, as well as economic growth in emerging markets, fueled the need for practical and decent business products that allow content to be translated into different language pairs.

Neural Machine Translation, Statistical Machine Translation, and a Little Bit of History

Neural Machine Translation (NMT) uses artificial intelligence to learn the rules of different languages and constantly improve. It works like a neuron that learns from specific materials and can predict the probability of a sequence of words.

Why is NMT so popular?

Improvements in learning algorithms, the ease of obtaining data to train the translation engine, as well as having the computational power necessary to train computers with a massive amount of information, have popularized NMT in recent years, to the point that is becoming a standard in MT, being adopted by different companies such as Google and Microsoft.

In many scenarios, NMT performs better, yields better results, and is much easier to maintain than a rule-based engine.

Statistical Machine Translation (SMT)

This model was promoted by IBM in the early 1990s. It evolved from word-level translation to phrase-based translation.

It’s training is based on creating a model that contains a sentence in a source language and its corresponding translation in the target language, creating a multilingual database.

Some MT advantages to think about…

• Some CAT (Computer-Assisted Translation) Tools allow the major MT providers to be integrated into the tool, either through a plugin or API (Application Programming Interface).
• Being able to translate many words, from many language pairs in a matter of minutes can drive down costs and increase delivery time.
• Machine translation is very fast. Like really fast. Thousands of words can be translated into multiple language pairs in a matter of minutes.
• In an MT workflow, the human translator does not disappear, but rather participates in the post-editing process, allowing the result obtained from the MT to be refined.

Conclusion:

Machine translation has come a long way, from the first experiments in 1946 to be able to translate a large volume of text in a matter of seconds using an engine that imitates human neurons.
Even so, MT is still in constant evolution, with improved algorithms and greater computational power.

GPI’s Machine Translation (MT) implementations ensure that NMT is a good candidate for the client’s needs. Firstly, carrying out a test project to determine the human translation effort required to edit the output, as well as the creation of a custom engine based on the content and the desired language pairs. GPI’s NMT solutions ensure savings and greater productivity.

The post Neural versus Phrase-Based Machine Translation appeared first on Globalization Partners International.

^{Article originally published on: https://slator.com/us-health-agency-mandate-machine-translation-post-editing-for-critical-text/} 

Despite vast improvements in some aspects of translations, many professionals across private and public sectors tend to agree that healthcare Machine Translation (MT) cannot fully replace human translation capabilities when it comes to critical medical subjects. This is evident in the recently proposed rule to Section 1557 of the Affordable Care Act published by the US Department of Health and Human Services (HHS) drawing the line between human versus machine translations.

To clarify the differences by definition, the rule states:

“We propose to define ‘‘machine translation’’ as automated translations, without the assistance of or review by a qualified human translator, that are text-based and provide instant translations between various languages, sometimes with an option for audio input or output. This is in contrast to human translation, which is context-based and captures the intended meaning of the source. This definition is based on literature addressing the use of machine translation in the clinical setting, which we believe captures the automated translations that are being used in the health care setting.”

During the pandemic, many healthcare patients with Limited English Proficiency (LEP) complained “because they were unable to sign up for Covid-19 vaccines on websites using machine translation or found translated information confusing because of inaccuracies in some translations”.

What’s more concerning is the high level of “inaccuracies when it comes to machine translation in the health care context” as the recent literature review pointed out. In fact, it was revealed that “all studies indicated error rates so high” that machine translation is found to be ‘‘unacceptable for actual deployment in health settings.’’

Therefore, the US Health Department recommends regulating healthcare machine translation output by requiring MT-translated materials to be reviewed by a “qualified human translator”, more importantly when it is “critical to the rights, benefits, or meaningful access of an LEP.”

The post Healthcare Machine Translation Needs Post-Editing appeared first on Globalization Partners International.

The effect of Artificial Intelligence (AI) on humans has increased over the last few years. From making our lives easier with online search recommendations, voice assistants and facial recognition logins, to facilitating advances in healthcare, identifying pandemics, and helping alleviate starvation. Artificial Intelligence is the ability of machines, especially computer systems, to simulate and perform human tasks and processes. Specific applications of AI include natural language processing, speech recognition, machine vision, and more.

In this blog, we will talk about AI and the benefits of using and implementing it in the localization industry.

Power of Artificial Intelligence

Artificial Intelligence offers several benefits making it an excellent tool that simulates human intelligence, including:

Reduce human errors: One of the advantages of AI is that it can significantly reduce errors and increase accuracy. The decisions taken by AI is decided according to information previously gathered and a certain set of algorithms. When programmed correctly, the errors can be reduced to zero.

No Risk: We can count on AI robots that are well programmed with a specific set of algorithms to perform risky tasks that humans cannot perform, such as defusing a bomb, diving into the deepest oceans and seas, flying to space, and much more.

Automation: AI can automate repetitive tasks without feeling fatigued, taking rests, or having breaks, as humans would need to.

Speed: AI systems can do tasks much faster than humans by finding and following specific patterns quickly. AI systems can analyze much larger datasets than humans and can perform complex mathematical calculations very quickly and accurately.

Artificial Intelligence and the Localization Industry

The localization industry is like many other industries that require human interactions and efforts. Artificial Intelligence can perform many localization-related tasks and processes faster, automatically, and with accuracy. Localization experts could utilize the use of AI in some areas in the industry such as Natural Language Processing (NLP), Machine Translation (MT), Terminology Mining, Speech to Text, and Text to Speech.

Natural Language Processing (NLP)

NLP is a subfield of AI that helps machines process and understand human language so that it can automatically perform repetitive tasks such as machine translation, spell check, text autocomplete, spam filters, etc.

One of the main reasons NLP is so important for businesses is that it can be used to analyze large volumes of text data, like customer support tickets, surveys, online reviews, news, social media comments, and much more. NLP can quickly help businesses analyze this data, provide analysis and reports, and enable decision-makers to make strategic decisions such as whether to localize a specific product, website, services manual, or marketing campaign for specific target markets.

Another usage of NLP in the localization industry is to analyze the source content to prepare a translation style guide and extract types of information for the linguists to process (i.e., person names, locations, organizations, etc.).

Machine Translation (MT)

Machine Translation is an automated translation performed by a well-educated computer installed with a machine translation engine. The MT engine provides text translations based on computer algorithms without human interaction. AI is used in the Machine Translation and Machine Learning subfields as the engine learns through high volume of source and target content. By using specific algorithms, the engine can provide the translations.

In the past, most MT products were based on algorithms that used statistical methods to try and provide the best possible translation for a given word. This technology is known as Statistical Machine Translation (SMT). SMT involves advanced statistical analysis to estimate the best possible translations for a word given the context of a few surrounding words. Recently, Neural Machine Translation (NMT) performs the process by attempting to model high-level abstractions into data, much closer to how it is undertaken by a human rather than the traditional statistical approach.

So, we can describe several types of Machine Translation as follows:

Rule-based machine translation (RBMT)

The earliest form of MT relies on linguistic rules and bilingual dictionaries for every language pair. A dictionary of the source language is used to select appropriate words in the target language. Syntax and grammar rules of both the source and target locale are observed, and the words taken from the dictionary are adapted appropriately (gender, grammar, word order, etc.).

Statistical Machine Translation (SMT)

In SMT, the MT engine uses statistical algorithms from analyzing existing human translations. It works by learning and comparing the source text with the model content. The translation is then generated based on the probability of occurrence in the target language. It works better for language pairs with similar word order.

Neural Machine Translation (NMT)

The NMT engine uses deep learning algorithms to train and educate itself and consistently improve. The engine functions similar to the human brain by using neural network models to create translation models.

Speech-to-Text

Speech to text using AI is a field in computer science that enables computers to recognize spoken language and transcribe it into text. Speech-to-text is different from voice recognition as the software is trained to recognize and understand spoken words and write the spoken text in the same language. Voice recognition systems focus on recognizing the voice patterns of individuals and taking an action (writing the text) according to the voice pattern.

Speech-to-Text can be used in several fields such as:

• Customer Service: Many agencies rely on chatbots that are based on AI systems to help answer customers’ questions. Since many users prefer voice chat, accurate and efficient Speech-to-Text software can improve online customer support services.
• Smartphone personal assistants, (Siri in IOS, Amazon Alexa, and Google Assistant), use AI-based Speech-to-Text systems to recognize the user’s speech and convert it to text, like creating a note, composing a text message, or searching the internet.
• Electronic Documentation: Many fields require live transcription of speech to be used for reference later or to generate reports. Some of these fields are remote meetings, medical services, online lectures, classes, and much more.

Conclusion

AI and its multiple fields and subdomains are increasingly being used in various industries and businesses to help improve and speed up repetitive processes. As the localization industry has many repetitive processes that take time, using AI has become a major demand in the industry. We can rely on AI to automate many human tasks to be done by computers with accuracy.

The post Benefits of Artificial Intelligence in the Localization Industry appeared first on Globalization Partners International.