Next-Generation Conversational AI for Database Access:
The Technology Behind AutoQL
How innovation in conversational AI is revolutionizing the way today's leading businesses access and leverage their data
Chapter 3
An Overview of Intent Classification
Behind the scenes of most modern chatbots is an AI intent classifier. Intent classifiers perform the function of recognizing intent in a user’s natural language (NL) question or statement — the thing the user wants to do or accomplish — and categorizing that intent in order to offer the best possible response.
AI intent classifiers analyze statements like “I want to buy a premium subscription, is it billed monthly or annually?” These systems leverage both natural language processing (NLP) and natural language understanding (NLU) to figure out that words like buy or subscription are likely to indicate that the intent of this message is purchase.
The AI-driven chatbot therefore needs established intent categories in order to classify intent. These categories must be tailored to the matter at hand. If the chatbot is being used for customer service at a SaaS company, intent categories might include needs help, demo request, downgrade, upgrade, and card expired. For a hotel booking chatbot, intents would be different and include things like book, cancel, change rooms, and change travel dates.
Intent classification is most helpful when there is a limited range of intents expressed by users who typically interact with the system.
When appropriate intents have been defined for a given chatbot, the AI must be trained to correctly match or associate those intents with a variety of different words a customer might use. This is where machine learning comes in.
By processing a large volume of example data — known as training data — the intent classifier begins to learn how to match human words to the intent categories it’s been programmed to know.
Intent classifiers can be used in tandem with other machine learning models that are equipped to understand some amount of context or make predictions about users’ needs in order to help facilitate an efficient and rewarding user experience.
But the goal of conversational AI is to close the gap between computers and humans, not to create a mediocre substitute for human-to-human interactions.
Considering that, intent classifiers still aren’t the best solution for facilitating flexible, dynamic user experiences that feel as intuitive as a conversation
with another person and yield the same results: namely, getting the exact information you’ve asked for instantly, easily, and consistently.
