Select Page

Natural Language Processing

How to talk with a machine?

 

Speech and image recognition software or programs capable of handling a conversation with a user were impossible to achieve 10 years ago. The first chatbots or text-reading systems were imperfect, made frequent mistakes, and did not recognize requests, while their responses were mechanical and expressionless.

Today’s cognitive software has made a giant leap in a very short time. With the use of advanced algorithms and the cloud’s computing power, fiction has become reality. Speech recognition has stormed out of the lab and into the commercial, financial and service markets. The coming years will bring about a revolution in human-machine interaction, for which we need to prepare today.

Artificial intelligence systems which are commonly used in business are in particular Natural Language Processing (NLP) systems and Voice User Interface in applications. What distinguishes them from other systems are cognitive algorithms, which have four characteristics:

Adaptation

 ability to “learn” changes and achieve better and better results.

Interaction

ability to interact with the user.

Memory

ability to “remember” the context of previous interactions.

Context awareness

ability to understand the meaning or context of the communication.

BlueSoft and Natural Language Processing

How to use NLP in business?

Many of today’s artificial intelligence systems may be implemented quickly thanks to the use of the cloud or languages and objects commonplace in the world of IT. Cognitive algorithms go hand in hand with Big Data and Cloud Computing, providing analyses and predictions which were impossible to achieve until recently due to the limited power of servers or limitations to data storage in relational databases.

More and more industries see the opportunities which are presented by speech recognition systems. Here are several practical applications of artificial intelligence:

 

  • Mobile banking – the future of mobile banking applications lies in recognizing users’ voice commands. Sending our a transfer, downloading a statement and sending it via e-mail, or checking a bank account balance by talking to an application are some of increasingly popular functions of mobile banking.
  • Transports, logistics and commerce – NLP enables monitoring the status of a shipment, canceling a flight booking or checking the availability of a product without engaging a consultant. An efficient call-center system, integrated with a speech recognition tool and sales or logistics systems, will answer customers’ questions just as well as a human employee.
  • Call-center – messages such as “for Polish, press 1” are a thing of the past. Today, the most advanced IVR systems can handle a real conversation with a user and, on the basis of their answers (and even their tone of voice), determine the customer’s need even before connecting them to a consultant.
  • eCommerce – speech recognition systems support inclusion of people with disabilities. Blind or otherwise disabled people will be able to order products using their voice, while an advanced image analysis will describe the product to them on the basis of an illustration on the website.
I have never, ever seen anything that was more intuitively dead-on to making the guest experience seamlessly delicious, effortlessly convenient, with the ability to talk to your room and say: ‘Alexa, I am here, open the curtains, lower the temperature, turn on the news.’ She becomes our butler at the service at each of our guests.

Steve Wynn, Chairman and CEO, Wynn Resorts, #248 Forbes 400

Amazon Alexa and its possibilities

BlueSoft’s experts have prepared a referential architecture presenting the possibilities of natural language processing. Amazon Alexa was selected for the case study due to its popularity and easy cloud integration.

By voice interaction with the user, our NLP system concept can:

  • order groceries (eCommerce based on SAP Hybris),

  • read a report on best-selling products in SAP Hybris,

  • read headlines on the Onet.pl news portal,

  • make a BLIK transfer

 

 

 

BeeBot

We estimate that over 50% of communication performed by consultants can be handed over to chatbots. That is why we decided to create our BeeBot:

Our innovative BeeBot can:

  • communicate via e-mail and messenger
  • be integrated with any system of the client

carry out tasks which were previously handled by employees – e.g. measure customer satisfaction, pitch the company, support recruitment processes, etc.

Thanks to BeeBot, employees can be delegated to more ambitious and development-oriented tasks, as it will relieve them of mechanical, repetitive work.

 

 

Lab classes, articles and trainings

BlueSoft has been organizing training and lab classes since 2015, and has participated in numerous conferences focusing on Machine Learning.