24Jun 2021

June 24th, 2021|Categories: News, Solutions|Tags: |

Not all datasets are created equal

A recent survey conducted by KPMG shows that organizations have sped up their adoption of Artificial Intelligence in response to the COVID-19 Pandemic, including 72% of industrial manufacturers, 57% of technology companies and 53% of retailers. As such, machine learning development services are beginning to enter traditional business structures like call centers, corporate news outlets, eCommerce and online learning platforms.

And while commercial applications of AI powered customer experience vary from speech and audio to video and image and even more common forms of digital content, the success of these initiatives rely on highly customized and trained learning engines.

We see many machine learning teams turning to off-the-shelf datasets in order to get their projects off the ground, which will without a doubt provide them with a quick, cost-effective solution and will for sure deliver initial results. Also, most organizations do not have access to a team of experts, including engineers and annotators, to deploy the necessary improvement models.

Yet, the positive outcome of any AI project hinges on the quality of training and testing data, whether your product is: a virtual assistant, a wearable, an autonomous vehicle, or any […]

14Feb 2020

February 14th, 2020|Categories: News|Tags: |

We Hear You – But do you understand?

Your voice. Unique to you. The words you use, the sounds you make when speaking, pronunciation, inflection, intensity. All uniquely you.

Siri, Alexa, Cortana, Bixby, Xfinity, BMW, Mercedes, Citicorp, Bank of America, and many more are incorporating voice technology into their products. From call centers to automobiles we can now talk to devices and hope they understand what we are saying and giving us responses that are useful.

We want devices to be smart, intuitive, and most of all, we want them to understand our unique voice. My voice devices need to understand how I structure my sentences, my mixture of slang picked up from countries around the world, choice of words and the meaning for me. Slang that I use because I picked it up in my travels and it fits my style. My voice device needs to understand me better than my therapist.

For devices to understand everyone, we need to train the engines and that requires data. And not “Speak to your machine” data, but human conversational data at thee most basic of levels. Human to Human. From Southern Drawl to latest urban slang, the classic New Yorker to […]

8Jul 2019

July 8th, 2019|Categories: News|Tags: |

Scott Schwalbach joining e2f as VP, Partnerships

SAN JOSE, California – July 8, 2019 – e2f announced today that industry veteran Scott Schwalbach has joined the company as of July 1st as VP, Partnerships.

After 17 years at Microsoft as International Program Manager, Scott has worked at SDL, Vistatec, Alpha CRC, Pactera and Cloudwords. For the last 9 years, he is also co-moderator of the successful Localization Unconference.

“I know and appreciate Scott for many years, and I’m thrilled to welcome him to our team,” said Michel Lopez, e2f’s CEO. “Scott is bringing a wealth of experience and knowledge that will help us reach the next level as we follow a path of rapid growth.”

Scott will be focusing on developing partnerships in the Translation, Localization and Data Annotation spaces.

From Translation to AI-Driven Language Experiences, e2f is driving the world of Voice amongst the Internet of Things. Based in Silicon Valley with global offices, we offer an array of services to meet your changing needs, exceed the highest in quality standards, all the while ensuring your customers receive the best in market language experience.  Our scalable, adaptable and driven approach to your needs is what makes e2f your partner of choice. […]

2Jun 2019

June 2nd, 2019|Categories: Localization Tips|Tags: |

The Importance of Context in Translation


Context please!

This is a seemingly straightforward request, variations of which are all too common among linguists—myself included. Though perhaps a bit curt, this request brings to light a frequent issue that all translators and reviewers face: the lack of context.

We often see ambiguous abbreviations, unclear phrasing, or even just have a single word, the meaning of which is unclear in isolation. For example, I remember coming across the word ‘Back’ while reviewing a gaming project. Lacking any context, the meaning was ambiguous. It had been translated in the sense of “go back” or “return”. However, after reaching out to the client to receive clarification, it turned out that it was intended to refer to the body part.

Thus, what might be crystal clear to the content’s author or to a client is not necessarily clear to its reader—or in our case, the translator. And such clarity cannot be expected: we cannot know the details of every product on the market, every event, every website.

Segmentation and Lack of Context

Another potential reason that lack of context is such a common issue is the way some modern translation or CAT (computer aided translation) tools work: the […]

28May 2019

May 28th, 2019|Categories: Localization Tips|Tags: |

Levels of formality across languages

Levels of formality are key to define the tone of a given speech. Various writing styles, idiomatic expressions, verbal morphology, and pronoun variations may have a different formality level across languages, countries, social groups, or professional domains. These variations may have an impact on those trying to immerse themselves in a different culture.

Across languages, levels of formality can largely be grouped into four categories: formal, neutral, informal, and vulgar.

How formality is represented in different languages

While some languages reflect their level of formality by merely employing a more formal or informal tone, others rely on specific grammar resources to avoid any kind of ambiguity (eg., variation of the second person pronoun “you”). In this article, we will use English, Spanish and French to explain and give some examples of how formality perception can vary depending on the language.

English, simple but effective

English is one of those languages that can solve its register needs without any specific grammar convention. “You” is the only second person subject pronoun in English language, and verb conjugations rarely vary across different subjects. Thus, formality entirely depends on other syntactic elements. The strategies to reinforce the idea of formal, neutral, informal […]