Understanding natural language is becoming an increasingly commonplace goal within the tech industry. Software built around these new technologies is changing our lives. These new systems are enabling our personal devices to better adapt and anticipate to what we need, right when we ask for it.

As a first step towards a complete, general linguistic framework, the declarative specification of an app must be designed as a functional model, which effectively realizes the mechanism of natural language communication. To ensure completeness, the model must take the language-based interaction between humans as its prototype. The model’s functionality and data coverage must be verified automatically by an efficient implementation as a running computer program. This combination of functionality, completeness, and verifiability constitutes the best scientific basis for the long-term success of upscaling the model.

Now, Alexandre Lebrun, the founder and CEO of VirtuOz, has launched Wit.ai, a developer platform and API that will let a software coder incorporate speech recognition and a natural language interface into any app or piece of hardware.

As per the developers site, Wit is a natural language for the Internet of Things (IoT) that enables developers to add a natural language interface to their app or device in minutes. It’s faster and more accurate than Siri, and requires no upfront investment, expertise, or training dataset. The framework works with mobile devices, home automation, wearable devices, robots, SMS remote control and anything else that is connected to internet.

photo credit: jamonation via photopin cc

How it works

Developers today face challenge in natural user interfaces. Mobile devices already have gesture and speech to interact with humans and similar natural ways to interact with technologies make it easier to learn how to operate them.

Wit contains an API based natural language platform, which can be used by developers to add modern a modern natural language interface to their app or device. The framework can be used to process natural language processing algorithms, configuration data, performance and tuning.

The platform gives a web console to configure what developers want Wit to do. The API is then call to send either text or audio natural language sentences and it get the structured information (JSON) in return. Developers can build a NLP analysis engine with Wit.ai by providing a few examples of the responses. Wit gets structured information in return and also learns from usage and helps improve the configuration as per usage.

The company says behind the scene, Wit combines various state-of-the-art Natural Language Processing techniques and several speech recognition engines in order to achieve low latency and high robustness to both surrounding noise and paraphrastic variations.

Wit determine the user intent and return the results back. As per Wit developers, intent is something the user wants to do such as ask about the weather at a certain place, ask about the result of a sports game, create a reminder in their agenda, and ask about their schedule for a certain day and more such actions.

Using machine learning, Wit.ai can accept the intent and intelligently combine the results as well as specialized vocabularies. For this purpose, the developers have incorporated several language processing engines that run in parallel, including the open source CMU Sphinx project developed at Carnegie Mellon.

“While natural language is hard, there is also a lot of overlap in what people want,” says Lebrun. “When each developer gives some examples, Wit understands and connects that example to others like it, and those new phrases become part of the bigger dataset.”

The concept is an instant hit among developers. TalkTo, the startup that provides free app to let users get answers from local stores without calling them, said Wit is a revolution in the voice and language industry. Riley Crane, co-founder and CTO of TalkTo said Wit’s uncanny ability to turn free form language into structured data has transformed TalkTo’s ability to get answers from businesses by helping us understand the meaning of each question. Their API provides a new and much needed superpower for developers by forming a bridge between the unstructured world that we humans live in and the rigid exacting world of machine code.

Future is on IoT

Cities themselves have sought solutions using connected devices and the Internet of things, such as can be seen in Kyt Dotson’s City of Paradigm: The Internet of Things article citing such innovations as the IBM smart cities initiative. However, programming projects such as Wit open up a different and equally important class of IoT connectivity, on a smaller scale, where humans interface with machines.

Google, Apple, IBM, AT&T and other enterprise are investing millions of dollars to master the natural language processing. Google has awarded over $1.2 million to support several natural language understanding research awards given to university research groups doing work in this area. Research topics range from semantic parsing to statistical models of life stories and novel compositional inference and representation approaches to modeling relations and events in the Knowledge Graph.

Wit had already been here for some times, and in that season has come to register more than 2,000 developers interested in the platform and already has major customers in the automotive world, hardware, television and various startups based in automating chores.

Lebrun himself declares that the platform is going to be a revolution in the industry of voice and language, and that one of its objectives is to try to break the manner in which speech recognition technology has developed in recent years.

The source code, SDK, and other information on Wit.ai is at GitHub and documentation can be found on Wit.ai site.