Got a question? Just ask Siri, Alexa or Cortana. Your answer will come immediately — and chances are it will be the right answer.
While these sorts of interactions may have once seemed like something out of the Star Tek franchise, where people and computers routinely carry on back-and-forth conversations, there’s nothing sci-fi or futuristic about any of this. Today’s digital assistants are entirely real and are enriching our lives in countless ways. For that, we can thank artificial intelligence (AI) systems and natural language processing (NLP) algorithms, along with the high performance computing (HPC) systems that make it all go.
So what is NLP? In a few words, natural language processing is a form of AI that allows a computer application to understand human language, either spoken or written. As Dell EMC data scientist Lucas Wilson explains, NLP applications “use computers to translate languages, convert voice to text and back again, and create human-like conversational agents to help customers deal with issues, questions and concerns.”1
Putting NLP to work
The use cases for natural language processing are all over the map, from automating customer service and help desk functions to analyzing and translating spoken or written language. Let’s look at a few examples of the ways in which organizations are putting NLP to work streamline processes, improve customer service and gain other business benefits.
In the retail world, AI-driven chatbots that leverage NLP are now just about everywhere — and they are multiplying rapidly. A recent study by Juniper Research found that the global number of successful retail chatbot interactions will reach 22 billion by 2023, up from an estimated 2.6 billion in 2019.2
For retailers, chatbots are now one of the keys to automating and streamlining customer interactions. They help shoppers get the information and answers they need quickly and efficiently. As Juniper Research notes, chatbots can help retailers deliver high-quality user experiences in a low-resource way, boosting customer retention and satisfaction, and reducing operating costs.
For healthcare providers, NLP systems can be one of the keys to automating burdensome processes, including the transcription of spoken or written notes from clinical staff members. NLP can also be used for “text mining,” or searching through documents to quickly find information related to patients and their care, the content of clinical studies and more.
As Gartner notes, NLP technology can turn text or audio speech into encoded, structured information that “may be used simply to classify a document, as in ‘this report describes a laparoscopic cholecystectomy,’ or it may be used to identify findings, procedures, medications, allergies and participants.”3
Chatbots are making widespread inroads into the banking and financial services industry. A study from Juniper Research found that the operational cost savings from using chatbots in banking will reach $7.3 billion globally by 2023, up from an estimated $209 million in 2019. This represents time saved for banks in 2023 of 862 million hours, equivalent to nearly half a million working years, the firm says.4
“Chatbots in banking allow heavily automated customer service, in a highly scalable way,” notes a Juniper Research author. “This type of deployment can be crucial in digital transformation, allowing established banks to better compete with challenger banks.”
The same NLP-driven technologies can be used to streamline and accelerate internal banking processes. For example, a Dell Technologies article notes that Lloyds Bank in the UK created a chatbot to help staff easily navigate the organization’s vast knowledge base.
Building and running NLP applications
While some NLP applications can require massive amounts of processing power from HPC systems, it doesn’t take a supercomputer to develop or run them. Many off-the-shelf HPC solutions are now available for training and running NLP applications. For example, the new Dell EMC Ready Solution for AI – Deep Learning with Intel delivers a ready-to-go solution for the development of AI-driven applications, including NLP systems. It provides an optimized solution stack that simplifies the entire workflow, including all the hardware, software and services needed to help organizations get AI solutions up and running quickly.
The backend development technologies for NLP applications are also becoming more accessible. That’s the case with the resources made available via the Intel AI Lab. In 2018, the lab introduced an open-source library for NLP developers, called NLP Architect. This resource, available through a GitHub repository, allows users to explore state-of-the-art deep learning topologies and techniques for NLP and natural language understanding (NLU), a closely related application. The NLP Architect provides an ideal platform for research and collaboration.5
Natural language processing is now just about everywhere, and it is helping organizations automate and streamline processes, improve customer service and reduce operational costs. And NLP systems are getting easier to build and deploy, thanks to new ready-to-deploy HPC systems that are optimized for AI applications and to new development resources like the NLP Architect from the Intel AI Lab.