it is a technology that involves Artificial Intelligence, voice recognition and messaging applications in order to interact with the user by simulating conversations very similar to human ones. If the interaction occurs via text, we can talk about Chatbot, if instead a spoken dialogue between man and machine occurs, we will talk about Voicebot.

Natural Language Processing; as the name suggests, it is a set of artificial intelligence algorithms capable of analyzing and understanding human language, mainly in the form of text.

Utterances, are the requests or phrases expressed by the user towards the bot. They trigger the start of the interaction between human and machine.
An utterance could be for example the phrase “is the new model of Teebok in white available?”

Intents are groups of concepts that indicate tasks and actions that the user wants to perform; they are necessary for the bot to interpret the utterances expressed by the human and therefore define the generic intention.
To define the intents, during the bot configuration phase, it is necessary to identify the main concepts that the machine will have to interpret, based on the application scope of the case.

If we wanted to structure the virtual assistant of an E-commerce for example, we should first establish the various areas of intentions within which the utterances expressed by E-commerce users can be divided. The expressions “I want to know how much the shoes cost” or “what is the price of the t-shirt?” will be traced back to the same intent that we could call “prices”; “how can I make a return?” “I would like to exchange a product” could instead fall within the intent “returns”.

Within a user-expressed utterance, entities are the words that define and specify intents, expressing a value or a variant connected to them and therefore defining them in more detail. Entities allow you to contextualize the user’s intentions.

Let’s take the example of our virtual assistant for E-commerce once again, within the same intent “purchases” we could find the sentences (utterances) “I want to buy shoes” and “I want to buy trousers”, where the first indicates the entity “footwear” and the second the entity “clothing”

Also known in Italian as Machine Learning, giving an exhaustive definition is not at all simple, its fields of application are vast and differentiated. We could however say that machine learning includes a series of different mechanisms that allow intelligent machines to increase their knowledge and performance, based on data acquired over time.

Thanks to feedback and corrections (carried out by a human operator in the case of supervised machine learning, or directly by the machine if we are talking about unsupervised ML) the system is able to learn how to perform functions and elaborate increasingly relevant responses.