Learning a new language can be difficult, but it’s worth the effort. The more you learn about the culture of your words and phrases, the easier it will be to understand how they are used in day-to-day life. What could you do with this knowledge? You might be able to better communicate with someone who speaks that language by using a common phrase or saying. Learning another language is also fun because it challenges your mind! No matter what age you are, there’s no reason not to try out a new tongue.

ML-based language acquisition apps – why they aren’t perfect

In the past, it was believed that machine learning (ML) would be a revolution in improving the quality of language acquisition applications for non-native speakers. Unfortunately, even when supported by ML technology, most learning mobile apps have serious shortcomings. Their functionality is somewhat limited, but technology is not the main culprit. Additionally, a drawback that is common and sometimes even a serious barrier is that usually machines, even those empowered with artificial intelligence and machine learning, cannot differentiate between the various levels of complexity that language entails. It is almost impossible for most applications to evaluate multiple “non-actual-text-related” factors behind what has been said and which are so important in language and speaking skills acquisition. This is where Voccent steps into the spotlight - an innovative solution that may change it all. One of its key features – prosody analysis technology that senses tone of voice, pauses and exclamations – may be a revolution in the industry. Voccent uses the most organic approach with natural human behavior in mind.

Current status of Language Learning

Have ever wondered the view rate between a social media content and the number of subscribers? On most of the channels it has some well predicted arithmetic, e.g. 1,000,000 subscribers would get you 500k views per each video or any other content. Though it’s fair for most of the type of content, the rules aren’t working out as good for Language Learning ones. You see, it’s different because a user needs to rewind and view samples again and again. Which, of course, generates a sheer part of view, hence monetization, for the Channel owners, yet, creates an enormous amount of inconvenience for the customers. Not only you’re not getting any positive/negative feedback from the red and black colors of the: “tubes”, “grams”, “toks” etc. You also generate a tremendous amount of useless traffic for the systems, consuming an unnecessary amount of energy. And the cherry for the cake to finish – hardly learning anything new. Why? You’re distracted most of the time, by the huge amount of video information coming in, or if it’s in your pocket by the need of app control and repetition into the void.

As usual, you’d expect we’re leaning to a solution and indeed there is. Voccent: Speaking and Listening learning platform. Curious? Let’s dive in.

Why Voccent was born

It all started a few years ago when serial entrepreneur and software engineer, Victor Brunko, realized that there was not a fully successful and efficient app for learning languages. He saw a gap in the market for a service that would improve language acquisition and advance accent training. After a while, Victor started working  on alternative  algorithms, despite loud opinions that ML technologies would be the future of language education. Victor’s experiences are coming from years of wide range achievements in such areas as: startups and team building, software engineering, mathematics, high-performance computing, multi-region big-data solutions and many others. His innovative idea was to use pure Digital Signal Processing, mathematics and performance computing to create a distinguishing solution for acquiring (not just learning) new languages via experiences based on psychology and neuroscience. It turned out that his approach was an excellent advancement.

What is so special about the app?

Currently, there are a wide variety of language learning apps on the market and all of them have their pros and cons. The team of engineers at Voccent being led by two founders: Victor Brunko and Lev Kaufman has made the most out of a non-ML approach, which arguably hasn’t been accomplished by others so far. The Voccent app utilizes: Listen-repeat cycles using real speech reference, Repeat-feedback cycles using real speech analysis and feedback, Non-language related speech analysis (the app is able to analyze audio-based sounds like people, animals, instruments, etc.) Of course, Victor and Lev did face some challenges while developing their platform. One could argue that the app is too sophisticated for a broad target audience, however, Voccent’s CEO highlights that the service has been designed for both average customers and average companies. It can be easily used by travelers, multilingual children, people who want to listen and speak without reading and writing, as well as call centers, speech therapists and companies training their employees. The platform may even have potential to be used for certain government services in the near future!

Where does the app go next?

To continue to make its mark on the industry, the Voccent platform needs even more innovation – more scientists, more engineers and more community drivers. Voccent CEO Victor Brunko envisions his language learning application having an impact in the industries of technology, travel, HR, education and many more. He dreams that Voccent will cover most (if not all) languages and regions, will connect more social groups, and will analyze even more features of communication and audio experiences. Will they succeed? The potential is definitely in the creator and technology, so only time will tell.