Skip to main content

AI Speaks Series Part I: How to derive business value from AI-based Natural Language Processing

Author by Swaminathan Venkatesh

We primarily interact with digital devices – a phone, a hand-held device or computer - with words, both written and spoken. A quick search, asking Siri a question, sharing our opinions, writing reviews or even registering our emotions through emoticons. All of these text-based interactions form a part of an exponentially growing source of human generated data that provides insights into the psyche of the customer. 

Businesses also deal with huge amounts of text data - in the form of digital records, internal reports, service logs and case files. Add to it customer interaction data like conversations with customer service or sales representatives or on social media – you are looking at a treasure trove of customer insight data that could help businesses make strategic decisions. This is where Natural Language Processing (NLP) enters the discussion to help you process this data. 

iStock-1225376238-(3).jpg

NLP is a field of artificial intelligence that deconstructs vast amounts of text data and codes the semantic relationships between words.  The goal is to create a system that can not only understand and quantify the intent hidden within human language, but also generate sensible text responses that mimics a human. NLP algorithms are powering innovation with ease while staying hidden behind everyday interactions in the digital world. Word suggestions that show up when you type into Google or your phone chat window and language translation apps are some examples.


Contrary to structured data that is clearly defined and organized in an easily searchable manner, text data is unstructured - lacking a format or organization. This is the main challenge confronting NLP researchers, who work to create algorithms and frameworks that would merge the field of linguistics with the power of computation. Current advances in NLP are a culmination of research effort spanning over 50 years. Here are some of the areas where businesses can take advantage of NLP:

  • Campaign personalization or tailoring advertisements to what a customer wants
  • Improving product sales, address concerns and reduce churn by analyzing customer feedback 
  • Creating AI solutions that can record, interpret and visualize trends based off digital records and emails
  • Customer facing chatbots to provide timely responses to customer inquiries 

Implementing NLP-based AI solutions helps businesses sift through constantly growing sources of text data to identify emerging trends, understand customer sentiment, uncover key ideas, and segment topics and customers. Through this "AI Speaks” blog series, I hope to introduce you to the nuances of how AI learns to understand and interpret the intent hidden behind human-generated text data. At the same time, I hope you will also get an idea as to how your business can improve products and services by leveraging NLP. In AI Speaks: Part II, I will address how NLP algorithms that understand math can make sense of the complexities and connotations of human words strung together in sentences.

Author

Swaminathan Venkatesh

Data Scientist