The Learning Curve, Part 5: Overcoming Multicultural and Multilingual Differences

Learning about Samsung Research centres across the globe, we delve into the world of Galaxy AI and its role in empowering users to unlock their full potential. Galaxy AI has expanded its language support to include 16 different languages. This means that more individuals can now enhance their language skills, even without an internet connection. The on-device translation feature, which includes Live Translate, Interpreter, Note Assist, and Browsing Assist, enables users to conveniently translate text in various situations. However, the process of AI language development may be of interest to you. During our recent trip to China, we had the opportunity to explore the significance of collaborating with fellow AI experts. In this particular instance, our focus shifts to Brazil as we delve into the intricate dynamics of cross-cultural and cross-border teamwork, all in the pursuit of making Galaxy AI accessible to a wider audience.

Brazil, a country with a population of over 203 million, is known for its rich cultural diversity and varied traditions. The official language spoken in Brazil is Brazilian Portuguese. Meanwhile, Latin American Spanish is used by 22 neighbouring countries.

Teaching Galaxy AI to discern and distinguish regional differences posed several challenges due to the intricate variations in both Brazilian Portuguese and Latin American Spanish, which are widely spoken. That’s why Samsung R&D Institute Brazil (SRBR) joined forces with Samsung experts from Mexico, along with science and technology institutes SiDi and Sidia, and assembled a talented and diverse team to take on the challenge.

Lower Barriers, Higher Understanding

The team extensively utilised numerous sources and employed a blend of machine learning and language processing tools to enhance the AI model’s ability to accurately identify and interpret speech, written texts, and regional nuances. However, the local terminology and references to well-known individuals, such as sports teams, celebrities, and bands, can differ significantly from one region to another. Furthermore, various words can be used to convey the same idea. Language models require localised data to develop a thorough grasp of various languages for translation purposes. However, these variations can pose challenges.

 

As an example, the word for swimming pool varies across different Spanish-speaking countries. In Mexico, it is referred to as “alberca,” while in Argentina, Paraguay, and Uruguay, it is called “pileta.” Meanwhile, in Colombia, Bolivia, and Venezuela, the term for swimming pool is “piscina,” which is also used in Brazil but with a slight tonal variation. Colombians use the word “chévere” to describe something cool, while Mexicans prefer to use the term “padre.”

The team successfully tackled the significant challenges in AI language understanding and learning by developing more extensive language models, enhancing processing tools, and fostering collaboration across different locations and time zones.

Adapting and testing the model involved careful consideration of local slang and various ways of speaking. This required a close collaboration between the SRBR quality assurance (QA) team and development teams, according to Mateus Pedroso, Senior Manager and Head of Software Quality Lab at SRBR. Due to the time zone differences between SRBR, the QA team in Mexico, and the management team in Korea, new communication channels and processes were established to ensure alignment of results and progress sharing. This diverse collaboration sparked a vibrant exchange of ideas and solutions for Galaxy AI.

 

Communicating Success

Samsung’s commitment to open collaboration was evident in this regional project, which embraced an iterative approach and harnessed cutting-edge technology on a global level. In order to overcome linguistic and cultural barriers, the SRBR team had to gather and handle large quantities of data, constantly enhancing and perfecting their audio and text sources.

The teams established specific areas of responsibility to maximise the utilisation of the diverse skill sets available in the company’s Latin American offices. The SRBR development team played a crucial role in the project, receiving guidance from Samsung’s headquarters and working on updates to enhance the AI model. They also conducted extensive tests for various use cases.

According to Leandro Flores de Moura, Software Development Manager at SiDi, the testing phase involved a lot of communication and collaboration with QA teams to enhance the user experience. Each adjustment made during this phase required additional testing and review. “The language capabilities of Galaxy AI owe their success to effective communication and collaboration, in addition to technical expertise,” remarks Nathan Castro, a QA Test Developer at SiDi.

A Roadmap for Culture

What makes Galaxy AI particularly interesting for everyone involved is the fact that this wasn’t merely a language project. To them, language is a cultural guide that provides valuable insight into people’s heritage and identity.

“For SiDi’s QA team, this was an endeavor that will change the world by enabling cultures to come together and overcome the difficulty of communicating in different languages,” adds Estefanía Castro Suárez, Test Developer at SiDi. “Knowing we were part of this fills us with pride and motivation.”

“The way the SRBR team collaborated exemplifies what Galaxy AI sets out to achieve—making the world a smaller place through communicating, sharing and interacting with people, even those who speak different languages,” concludes Pedroso. “This capability will only grow as more languages come on board with Galaxy AI.”

Join Galaxy AI-Volution Squad Today!

Purchase our latest Galaxy innovation 

To learn more about Galaxy AI, visit: https://www.samsung.com/my/galaxy-ai

Leave a Reply

Your email address will not be published. Required fields are marked *