UZH and ZHAW Offer Certificate in Advanced Studies in Translation Technology and AI - Slator - Translation

Professional translation has undergone profound changes with recent advances in machine translation. Humans and computers are interacting more and more in the translation process. Translators must, therefore, be as familiar as possible with the functionality, possibilities, and limitations of computer-assisted translation.

To address this need, the University of Zurich (UZH) Department of Computational Linguistics and the Zurich University of Applied Sciences (ZHAW) offer a Certificate in Advanced Studies (CAS) in Translation Technology and AI.

The goal of this CAS is to familiarize translators with modern translation tools — in particular, neural machine translation — through clear examples and practical exercises.

Advertisement

The course is designed for translators who want to enjoy the benefits of life-long learning. One clear benefit of the program is that translators can gain an edge in knowledge for future professional practice. The course also enables translators to keep abreast of current developments in translation technology and relevant research findings.

Those with a master’s degree in applied linguistics (or a linguistic subject) and prior professional experience are welcome to apply. Individuals with a bachelor’s degree and professional experience may also be admitted upon review of their application materials.

The CAS in Translation Technology and AI will be conducted in German, except for one course day in the second module, “Spoken Language Processing: Speech Recognition, Speaker Recognition, and Automatic Interpreting,” which will be delivered in English.

Course fee is CHF 6,500 (USD 7,200) and classes are held on Friday and Saturday and will run from October 2021 to March 2022. Registration is currently open.

Send an email to weiterbildung@ifi.uzh.ch to express your interest. Deadline for registration is September 20, 2021.

The CAS in Translation Technology and AI graduated its first batch in 2020!

The UZH and ZHAW (Zurich University of Applied Sciences) community congratulates the 2020 graduates who were recently awarded a CAS in Translation Technology and AI. We celebrate their extraordinary achievement and wish them all the success in their professional future!

Course Structure

The CAS in Translation Technology and AI comprises 11 course days of 8 hours each, and it is divided into two modules.

Module 1 covers “Computer-Aided Translation and Basic Programming” and includes an introduction to the Python programming language on every morning of the first module. Learning the basics of Python will enable students to write small programs as well as strengthen their understanding of algorithmic language processing.

Module 2, “Machine Translation and AI,” begins with an introduction to machine learning and neural networks. Click on each Course Day here to read a description of what will be covered.

Background reading and course projects across the twin modules will help students develop the skills to adopt methods currently in use in today’s professional environment.

Assessment

Students are required to complete one certificate of performance per module. Module 1 must be successfully completed with a written test.

After Module 2, students will be asked to write a paper on a topic of their choice from Module 1 or 2. Participants should decide on a topic during the course of the module in consultation with a lecturer.

The paper must have a maximum length of 20 pages and should be written in either German or English.

ECTS Requirement

A total of 12 ECTS (European Credit Transfer and Accumulation System) credits are required to successfully complete the CAS in Translation Technology and AI.

Each module is equivalent to 6 ECTS credits. One ECTS credit corresponds to a workload of 30 hours. This consists of attendance time during the course days as well as preparation and follow up.

Upon course completion, the student will receive a Certificate of Advanced Studies from UZH in Translation Technology and AI (CAS University of Zurich).

For more information about the CAS in Translation Technology and AI and the next course, which starts in fall 2021, click here.

The First 'Google Translate' for Elephants Debuts - Scientific American - Translation

When a male African savanna elephant folds his ears while simultaneously waving them, he’s ready for a fight. When a female folds her ears and accompanies the action with an ear flap, that means she’s also issuing a serious threat. But when elephants come together and fold their ears while also rapidly flapping them, the animals are expressing a warm, affiliative greeting that is part of their bonding ceremonies.

Elephants possess an incredibly rich repertoire of communication techniques, including hundreds of calls and gestures that convey specific meanings and can change depending on the context. Different elephant populations also exhibit culturally learned behaviors unique to their specific group. Elephant behaviors are so complex, in fact, that even scientists may struggle to keep up with them all. Now, to get the animals and researchers on the same page, a renowned biologist who has been studying endangered savanna elephants for nearly 50 years has co-developed a digital elephant ethogram, a repository of everything known about their behavior and communication.

“Without a multimedia approach, I see it as impossible to properly show and explain the behavior of a species, and we hope this will inspire other scientists to take a similar approach for other species,” says Joyce Poole, co-founder and scientific director of ElephantVoices, a nonprofit science and conservation organization, and co-creator of the new ethogram. “At a time when biodiversity is plummeting and the lives of elephants are being heavily impacted by humans, we also want to spell out to the world what we stand to lose.”

Poole built the easily searchable public database with her husband and research partner Petter Granli after they came to realize that scientific papers alone would no longer cut it for cataloging the discoveries they and others were making. The Elephant Ethogram currently includes more than 500 behaviors depicted through nearly 3,000 annotated videos, photographs and audio files. The entries encompass the majority, if not all, of typical elephant behaviors, which Poole and Granli gleaned from more than 100 references spanning more than 100 years, with the oldest records dating back to 1907. About half of the described behaviors came from the two investigators’ own studies and observations, while the rest came from around seven other leading savanna elephant research teams.

While the ethogram is primarily driven by Poole and Granli’s observations, “there are very few, if any, examples of behaviors described in the literature that we have not seen ourselves,” Poole points out. The project is also just beginning, she adds, because it is meant to be a living catalog that scientists actively contribute to as new findings come in.

“We know elephants behave and communicate with each other in complex ways. But until now, we have barely scratched the surface of just how complex that behavior and communication is,” says Lucy Bates, a visiting research fellow specializing in elephant cognition at the University of Sussex in England, who was not involved in creating the ethogram. “Now we have this foundation—freely available in the public domain—from which we can build a much more comprehensive picture of what elephants are doing and why.”

Ethograms are compilations of animal activities and behaviors, either in a specific context or for a species overall. Researchers use ethograms for studying behavior and making comparisons among ages, sexes, families, populations or different species. While a digital ethogram exists for laboratory mice, and another has been published in written form for chimpanzees, Poole and Granli believe the exhaustive, digitized Elephant Ethogram is the first of its kind for any nonhuman wild animal. The multimedia-based nature of the project is important, Poole adds, because with descriptions based only on the written word, audio files or photographs, “it is hard to show the often subtle differences in movement that differentiate one behavior from another.”

When Poole began studying elephants in 1975, scientists knew very little about their behavior. Her early research focused on musth, a periodic reproductive condition in male elephants that is characterized by testosterone surges and heightened aggression. Poole noted that the animals waved their ears as a threat, and she sometimes also noticed a low, pulsating sound accompanying this movement. At first, she thought it was produced by the ears whooshing through the air, but she soon realized that the sound was a vocalization. She began to wonder if the animals were making other sounds too low for her ears to fully detect.

To follow up on these observations, Poole teamed up with acoustic biologist Katy Payne. Together, they revealed that the different rumbles elephants produce contain some frequencies below the level of human hearing and that some of these sounds are so powerful they may be heard by other elephants miles away. This discovery helped solve a number of elephant mysteries, including how members of a family are able to quickly find one another after they split up and how the animals, without seeming to make a sound, are able to act in complete unison when a threat is detected.

“For a long time, people talked about elephant ESP,” Poole says. “Some vocalizations they make are so powerful and are transmitted through the ground as vibrations, acting like a kind of bush telegraph for elephants saying, ‘There’s trouble.’”

As Poole continued to study elephants, she realized that the meaning of many of the behaviors she was documenting also changed depending on the context. Tail swatting, for example, is usually used by one elephant to tell another to back off because it is standing too close. But mothers also use tail swatting to keep tabs on infants standing behind them. In still other contexts, elephants use it to pester or nag for attention.

For now, the majority of entries in the Elephant Ethogram come from three locations: Maasai Mara National Reserve, or the Mara, and Amboseli National Park in Kenya and Gorongosa National Park in Mozambique. The limited number of sites do not necessarily affect the breadth of behaviors in the database, however, because most elephant behaviors are conserved across populations. But the frequency of certain behaviors may differ depending on the study site. Elephants at Amboseli, for example, never dig for minerals because the soil there is salty, and Poole has only once seen an Amboseli elephant shake a tree to knock down seed pods because there are very few trees in the area. In the Mara, on the other hand, elephants often dig for minerals. And in Gorongosa, they frequently shake trees for seeds.

Stark cultural differences can also exist among populations. During the Mozambican Civil War, 90 percent of the elephants in Gorongosa were killed for their ivory and meat. Nearly 30 years later, the elephants there still act fearfully and aggressively toward people. “Their behavior is very different from Amboseli and the Mara, where we see almost no defensive behaviors toward people,” Poole says. “As Gorongosa experiences a rebirth under new protection and restoration, how long will it take for elephants to abandon those traditions?”

The answer will likely someday be cataloged in the Elephant Ethogram. Now that the project is online, Poole hopes other researchers will begin contributing their own observations and discoveries, broadening the database to include cultural findings from additional savanna elephant populations and unusual behaviors Poole and Granli might have missed. Wildlife photographer Kelly Fogel, for example, recently submitted rare footage of an elephant eating her own placenta after giving birth. And Elephant Aware, a nonprofit conservation group, sent in a similarly uncommon video of a calf trying to suckle from her dead mother. “Now that the Elephant Ethogram is publicly available, we hope that more of our colleagues will share unusual footage from their populations,” Poole says.

Already, though, the ethogram is an “invaluable tool” for young scientists who are interested in conducting field research on elephants, says Michael Pardo, a postdoctoral researcher studying African elephant vocal communication at Colorado State University. The ethogram will also help ensure that scientists are talking about the same thing when referring to a specific behavior, he says. “This is really important, because all too often in the scientific literature, confusion about terminology and definitions can result in researchers talking past one another,” Pardo adds.

Daniela Hedwig, a research associate at the Elephant Listening Project at Cornell University, agrees that the Elephant Ethogram is “a monumental achievement.” Hedwig studies communication in forest elephants in other parts of Africa, and the savanna elephant database will be “an extremely useful repository to draw comparisons between the two species,” she says.

Cynthia Moss, director of the Amboseli Trust for Elephants in Kenya, says the ethogram will also be particularly helpful in assessing the lives of captive elephants. Elephants held at circuses, zoos and work environments are deprived of intricate social connections and the ability to travel and to interact with a diverse environment, Poole says. This causes them to suffer from boredom and to develop stereotypic behaviors, such as rocking back and forth. The Elephant Ethogram can shed light on the severity of behavior differences between captive and wild animals, Poole says, and strengthen the case for ending elephant captivity.

Moss adds that the database will also be a valuable tool for wildlife managers and conservationists who seek to differentiate between natural, healthy elephant behaviors and ones induced by stressful conditions, such as poaching and habitat loss. The need for such comparisons is growing as elephants in the wild face mounting pressure to tailor their behaviors to a people-dominated world. “Elephants’ ability to culturally adapt is going to be critical for their future survival,” Poole says. “As they are forced to change, we’re going to learn a lot more about their creativity and flexibility.”

Need a cyber dictionary? Welcome to Grammar 2.0 - Sydney Morning Herald - Dictionary

“Can a dog be twins?” So reads the meme, a tweet composed by the nameless heroine, five words that soon go viral. Global. “The post had recently reached the stage of penetration where teens posted the cry-face emoji at her. They were in high school. They were going to remember ‘Can a dog be twins?’ instead of the date of The Treaty of Versailles, which, let’s face it, she didn’t know either.”

Some of the new emojis available. Credit:Unicode Consortium

The heroine hails from Patricia Lockwood’s novel, No One Is Talking About This (Bloomsbury Circus, 2021). A maverick mosaic, the book shadows our vaunted ‘poster girl’ through real life and ‘portal’ life, from bedroom to chatroom, all distinctions blurring into a single on-life of likers and followers.

Lockwood, an American poet, has a rare gift for phrasing. Cancel culture gains real lustre in her hands: “Every day their attention must turn, like the shine on a school of fish, all at once, towards a new person to hate.”

Likewise, the gnaw of web addiction: “this metastasis of the word next, the word more.”

Her humour is equally wild. Take the emoji-blindness of her mother, who sends recipe texts that read like porn. The daughter has to pounce. “NEVER SEND ME THE EGGPLANT AGAIN MOM, she texted. I DON’T CARE WHAT YOU’RE COOKING FOR DINNER!”

Any book dealing with the net is about communication. More than Beowulf or the telephone, the web is reshaping English. Gretchen McCulloch, a Montreal linguist and author of Because Internet (Riverhead, 2019), attributes the impact to “weak ties”, where a tribe’s cohort of adjacent outsiders relay new grammar to distant platforms.

Loading

That’s the net, in a nutshell. Each post we send, each quip we share, is a raindrop striking the big pond’s surface, dimpling and radiating from my space to yours. At one point in Lockwood’s novel, the heroine is feted at a literary festival, thanks to her “doggerel”. She shares a panel debating which is funnier: sneezing or sneazing.

Tellingly, even in chatrooms, such slang-centred discussions seldom occur, usurped by adoption instead. As Lockwood writes: “Spellings of the word baby that the portal had lately cycled through: babey, babby, bhabie. Middle English had seen similar transformations: babe, babe, babi. Yet in every variation, the meaning shone through, as durable as a soul, wrapped in swaddling clothes.”

Vivaldi adds mail, calendar, RSS and translation tools to its privacy-focused browser - Engadget - Translation

Vivaldi has released a major update for its eponymous web browser for privacy-minded power users. Version 4.0 brings with it a translation tool, along with beta versions of Vivaldi Mail, Calendar, and Feed Reader. The update is available now on Windows, Mac and Linux and Android devices. 

Vivaldi built its translation feature into its browser. The tool is powered by Lingvanex, a Cyprus-based company that makes translator's for a wide range of platforms including voice calls and smartwatches. As part of its focus on privacy, Vivaldi says that all your translation activity will be kept away from third-parties on its servers in Iceland. 

With the update, a new translate icon will appear in the desktop browser's address field when a web page loads. You'll be able to translate a full page to your default chosen language or use the drop-down menu to choose from other languages and additional options, including "never translate" and disable auto-translate. The feature is also available on Vivaldi's mobile browser via the menu.

In terms of the new beta features, Vivaldi Mail is arguably the biggest addition. The client can show you multiple accounts in a single inbox and works with providers that support IMAP and POP3, including Gmail. Vivaldi says it will do "the heavy lifting for you" by detecting mailing lists and automatically categorizing your mail to make it easier to find. You can also tweak the three-panel layout to a horizontal split view.

Vivaldi's Feed Reader, meanwhile, lets you follow RSS feeds, YouTube channels and podcasts with no "promoted content" or "spying." Finally, Vivaldi's Calendar is its alternative to Google and Microsoft's offerings. The new tool supports several online calendars like Fastmail, Zimbra and iCloud. You can also choose between three different views including minimal, full view and compact.

With so many new additions, Vivaldi is aware that some users may not want such a busy browser experience. So, it's debuting three layouts: Essential, which keeps things minimal; Classic, complete with panels and the status bar; and Fully Loaded for the complete range of features including its beta tools.

All products recommended by Engadget are selected by our editorial team, independent of our parent company. Some of our stories include affiliate links. If you buy something through one of these links, we may earn an affiliate commission.

Mount Sinai eases translation in 200 languages through its Epic telehealth platform - Healthcare IT News - Translation

When telehealth visits increased exponentially during the pandemic, patients at Mount Sinai Health System with limited English proficiency, or LEP, were at a great disadvantage to receive equal access to services because of the language barrier.

THE PROBLEM

Telehealth platforms were not originally set up with LEP patients in mind. And Mount Sinai's patient population in New York City is very diverse – the health system needs to be prepared to deliver healthcare in more than 200 languages.

"Our pool of in-house interpreters is limited; it only covers a few languages," said Silvina de la Iglesia, associate director of language assistance at The Mount Sinai Hospital. "As a health system, we rely primarily on phone and video interpreter vendors to cover the demand.

"One of the most challenging problems we faced was not having interpretation services integrated into the different telehealth platforms that were being used," she continued. 

"Each platform offered different capabilities for three-way communication, and we needed to explore which one would work best for our patients. We chose to tackle Epic/MyChart in the first place and work in partnership with our language vendor, Language Line Solutions."

On this platform, an invitation to join the encounter could only be pushed via email or text message in real time. This was possible when aligning the encounter with a staff interpreter, but not with the organization's phone or video vendors, as those interpreters operate in an enclosed environment on an on-demand basis.

"Providers were initially calling Language Line and placing the interpreter on speakerphone or holding cell phones in their hands and hoping that the video platform would pick up the sound," de la Iglesia recalled. "This was a very frustrating experience for providers and patients as the audio quality was poor, with lots of residual noise and echo on the line."

"And as with all technology and use-cases, rarely is there a one-size-fits-all solution. It is important, whether interpreters or caregivers, to have as many options as possible for flexible technology to allow another party to join."

Natalya Sholomyansky, Mount Sinai Health System

Video visits are heavily utilized at Mount Sinai and a great option for patients who do not want to come into the office; they can be accessed via a mobile app or on a computer with a webcam. Neither needed the ability for the patient or other individuals joining the video visit to be called via phone number since this was accessed via login credentials.

"Our in-house interpreters could be invited to and join a video interaction with our patients using the existing platform," said Natalya Sholomyansky, director, IT digital health, at Mount Sinai. 

"However, external vendors that supplement cannot leverage this workflow or join video visits. Providers can only access those external Language Line partners by dialing out to an 800 number, a capability that did not exist in the video visit platform. This left only the in-house language interpreters able to support patients who needed them during the video visit."

PROPOSAL

Mount Sinai turned to health IT vendor Caregility for help. Caregility offered the option for the provider to dial a phone number to Language Line and have the external vendor join the video visit with audio.

"Our providers and non-English-speaking patients needed instant, seamless access to language interpreters to communicate effectively," de la Iglesia said. "Language Line has a pool of 14,000 interpreters in more than 200 languages operating 24/7. Caregility offered the option for the provider initiating the encounter on the telehealth platform to dial out to a toll-free number to connect with Language Line within seconds."

Mount Sinai leveraged this capability by building a call-routing architecture behind the scenes. With the help of Language Line, the provider organization created dedicated toll-free telephone lines for the top five languages – Spanish, Mandarin, Cantonese, Russian and Bengali – and another one for the remaining languages. The dedicated telephone lines are answered directly by interpreters, reducing the connectivity time to less than 10 seconds.

MEETING THE CHALLENGE

This new solution went live on October 26, 2020, with the more robust integration features. Mount Sinai expanded utilization 100-fold in three days.

"When the provider is ready to start a video visit, they can have an external interpreter join," Sholomyansky explained. "They do so by selecting one of our language lines and the language that is needed. The interpreter will then join the visit by audio only."

The IT teams from Mount Sinai, Language Line and Caregility worked together to place the patient at the center of this initiative and find solutions to address the language and technology barriers, de la Iglesia said. The focus was to find a solution for Epic/MyChart, the telehealth platform most widely used by the health system for ambulatory services.

"We achieved integration of interpretation service into the platform via audio in a very short period of time and with minimal testing," she noted. "The sound quality has greatly improved as it comes from a single source.

"The most important feature to highlight is that it is provider- and patient-friendly," she added. 

"The provider simply clicks on the Add Interpreter button and selects a desired language. The patient can receive the link on any mobile device and simply click on Begin Encounter to join. Once on the encounter, the patient can see the provider on the screen but only hear the interpreter, which fosters a more direct patient-provider rapport supported by the invisibility of the interpreter."

RESULTS

From the time Mount Sinai implemented this solution to date, more than 600 telemedicine encounters have taken place using the Add Interpreter button. To ensure the health system was going in the right direction, it surveyed non-English-speaking patients and obtained feedback from providers using the service.

"Some quality metrics that indicated success were connectivity time, audio quality, ability to understand the interpreter and availability of interpreters," de la Iglesia reported. 

"We also identified some opportunities for improvement, like educating providers to add the interpreter before the patient joins an encounter. Patients reported they feel more welcome when they are greeted in their preferred language. A total of 88% of surveyed patients would recommend telehealth services to family and friends."

ADVICE FOR OTHERS

The ability to have a third party join a telehealth session in general is an important one, Sholomyansky said.

"And as with all technology and use-cases, rarely is there a one-size-fits-all solution," she noted. "It is important, whether interpreters or caregivers, to have as many options as possible for flexible technology to allow another party to join."

From the language access and patient experience perspective, de la Iglesia would encourage stakeholders to work together, in a creative and agile manner, and explore all possibilities the telehealth platform offers. Oftentimes, it does not require a lot of financial investment or bringing a completely new vendor into the picture, she said.

"Knowing the different communication needs of the patient population your hospital or health system serves also will provide great guidance to build a roadmap," she said. "When we first launched this technology solution as audio, we assumed our patients expected to see the interpreters on video; however, 86% of the surveyed patients indicated satisfaction with audio interpreting as it is.

"Providers also reported that seeing an interpreter on the screen could be very distracting," she concluded. "Video is imperative for patients who are deaf and/or hard of hearing, but not for spoken languages. Patients are driving the way we deliver healthcare, and we need to keep them engaged when implementing solutions."

Twitter: @SiwickiHealthIT
Email the writer: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.

Volunteer Center offering Spanish translation | Faith & Community | journaltimes.com - Journal Times - Translation

RACINE — The Volunteer Center of Racine Inc. will be offering Spanish translation services to clients of the Galaxy Digital volunteer database on volunteerracine.com. The database is a cooperative venture with United Way of Racine County.

Samantha Mayer, a Spanish and political science major with a minor in economics, will provide the translation services as part of her senior graduation project at Carroll University.

Mayer will be available at the Volunteer Center before noon, Monday, Tuesday and Wednesday through November.

For more information, contact Michelle Ortwein, executive director at 262-886-9612 or mortwein@volunteerracine.org.

0 Comments

Get local news delivered to your inbox!

The Age Of Machine Translation - CIOReview - Translation

Neil Currie, Head of Data Management, Crowe

With the entrance of Covid on to the world stage, employees and companies solved a problem that many had been working on for over a decade, the ability to make remote working a reality. Overnight, millions of people were told to work from home, and infrastructure transitioned to support virtual desktops, and meetings via Video conference became normal. Companies like Zoom saw their value increase in multiples, and the beginning of a new operational model appeared. Of lesser note was the increased adoption of some of the other tools necessary to support a new globalised workforce where many of the traditional boundaries had been removed.

These included translation engines, many of which had been operating on a smaller scale in the background but were now becoming part of everyday business life. As we know, adoption drives development, and we are now starting to see the much-needed greater sophistication necessary to support machine based translation. This is a realm where nuances, context, and ontologies matter. To provide a simple example, if you saw the acronym “HA” in the text, context can become very important, for a family doctor, it might refer to headache, a cardiologist heart attack, and endocrinologist hepatitis A. If you saw the word “body” in the text, for the medical fraternity, it might refer to a person, to the corporate world, a corporate entity, and to an association its membership. The translation of the word is impacted by the surrounding text and the document origination. Translation done manually will see multiple nuances based on the knowledge of the person, their experience, and interpretation of what is meant limited by their level of understanding. In the world of machine-based translation, we see a world without these nuances where the translation is done based on dictionaries tailored to different contexts and environments. Embedded images remaining in the original format and non-machine readable documents fail translation. None of these issues is a surprise to the people who use these technologies to support their products, but to the information consumer, they remain a source of frustration where the recipient is expecting to see a perfect translation fit for their use. Fed by the “On-Demand” requirements of how our interactions have evolved, the need to refine the capabilities for this secondary level of background technologies is growing and in the same way, we have seen an industry sector develop to provide training datasets to support machine learning it is likely we will see one for translation dictionaries to support this growing need.

We see today ready adoption of the current flawed model in places where accuracy is less important, for example, on-demand translation on websites, instruction books. In some roles, it is unclear when adoption will become mainstream, these include the law profession, general medical profession, and those where a greater level of accuracy is important. For those areas like academic papers, scientific journals, and those items requiring peer review and opinion, it is likely they will remain manually translated because of the requirement for nuances, interpretation, and environmental distinctions to support the hypothesis proposed and review by peers.

Today Alexa, Google, and Siri’s voice technology is already a part of our life, and the AI-supported technologies that allow these to operate in multiple languages with tonal nuances of people from different areas can hopefully be used to support the contextual interpretation, nuance etc. required to mature document-based translation as these technologies continue to develop.