While globalization and technology make it easier for the population to communicate real-time, there are still hurdles in fully understanding each other. It becomes frustrating when a website content cannot be translated into one’s language or even into the widely used language of English. Or sometimes, the translation is problematic or even comical.
Andovar, a provider of multilingual content solutions, aims to use language technologies to help businesses reach a market that has otherwise been neglected in the past. With the mantra of glocalization, the company hopes “to offer all solutions under one roof, including translation, desktop publishing, testing, audio, video, subtitling, engineering, language technology advisory, testing, and more. This combined with production in low-cost locations is what gives us a competitive advantage over others.”
“Language technologies are an ever-growing offering and their role is expanding all the time,” said Conor Bracken, CEO of Andovar. “Historically, the first ‘technologies’ used in translation were typewriters or computer word editors. Then, with the advent of Computer-Aided Translation Tools and Translation Memory, the daily work of a translator changed immeasurably. Since then, terminology management, workflow automation, and Machine Translation are widening their adoption and help increase the speed and minimize overheads.”
Andovar was founded in 2007 in Southeast Asia, with Western ownership and management. The company focuses on emerging markets and technologies, such as cloud software, gaming, mobile, and websites. Its main office is in Singapore and maintains offices in Colombia, India, Thailand, the United States.
Bracken said that the company is focused on emerging markets because these are the new frontier in localization and where others struggle with complex scripts and lack of standards. Its main services range from text translation and content creation, through audio and video recording, to turnkey localization of websites, software, eLearning, video and games.
Andovar uses machine translation to address language challenges for an ever-growing global audience. Technology companies have been offering free internet to countries or locations that could be hardly reached by these services but most of the content is not in the language they speak, which somehow defeats the purpose.
Bracken said the use of machine translation is “very inexpensive and can easily scale to almost any volume. This has enabled the translation of a lot of content that would otherwise stay in the source language forever. One of the unexpected beneficiaries is minor languages or the emerging markets, which are not important enough to merit paying for human translation.” It also helps address the challenge of having a limited number of qualified translators as well as the cost of charging per word.
While Andovar helps companies to “glocalize” their content, Bracken is quick to point out that “translation is only one part of it. It is the most important part but would not be enough without other peripheral services.”
“Let’s consider some common formats of content: movies, video games, websites,” Bracken said. “In all cases, there is text to translate, but in addition to translation, one must also extract and then re-integrate the text. Movies require recording of audio, or creation and formatting of subtitles. In video games and websites, the text is mixed with programming code and it takes expertise to separate the two. Once a game or a website is translated, the layout and user interface may have to be adjusted to that the new text fits and displays properly. This is done by another team of specialist.”
Bracken also explained that customers want content that seems especially created for them and not the word-for-word type of translation where the real meaning can sometimes get lost.
“This calls for input from culture and marketing experts, who knows how to reach the new target demographic,” he said.
The use of machine translation technology aims to expedite businesses localization process. It removes tedious human interventions and processes that could take up to days depending on the response rate. With Andovar, work is done in the cloud where all parties can collaborate. Once the material has been uploaded for translation, the content is divided among translators and revisions and comments are done within that working group wherever each member is located in the world.
“Such as system saves time, money, effort, and allows for 24/7 response time,” Bracken said. “Currently, such a workflow can be implemented partially or fully, depending on the circumstances.”
Machine translation has been in use for decades now but Bracken said its evolution has accelerated recently.
“Although the quality was often lacking, it became immensely popular and brought MT into the limelight again,” he said. “Other internet giants presented similar services soon after, the most well-known of which is now Google Translate.”
But much of its use is on the consumer side and companies either overlooked or neglected its advantages for commercial purposes and how it can boost business globally. The few translation companies that ventured into MT for commercial purposes took the chance to expand their specialization and improved their services using machine translation.
“In the last 5 years, the biggest advancement has been Neural Machine Translation (NMT), which is based on the paradigm of machine learning and is the newest approach to MT,” Bracken said. “NMT uses neural networks that consist of nodes conceptually modeled after the human brain. The nodes can hold single words, phrases, or longer segments and relate to each other in a web of complex relationships based on bilingual texts used to train the system. The complex and dynamic nature of such networks allows the formation of significantly more educated guesses about the context and therefore the meaning of any word to be translated. NMT systems continuously learn and adjust to provide the best output and require a lot of processing power. This is why this approach has only become viable in recent years.”
The interesting use cases of machine learning include sentiment analysis in any language allows to quickly identify whether the online chatter is positive or negative, and to recognize abusive behavior regardless of the language. It can automatically decide whether the content is suitable for MT, which applies a set of parameters to decide if the expected quality of machine translation. This is perhaps one of the most popular use cases, the subtitling of movies. In ideal circumstances, the transcription, time-coding, machine translation, creation of subtitles in target language and integration with the movie can be done with limited human input.
Bracken said Andovar plans to continue with the business model as “it’s been successful. We are always looking to expand production in low-cost countries to maintain our competitive edge. We also want to integrate technology and MT more by following latest developments to save time and money for our clients. In the long term, we would like to break into new markets, grow our customer base as well as average deal size.”
Image from Pixabay