If you’re starting a chatbot development project, and do not like it when technical people use technical terms to look smart, this article is a great start. We will explain the technical terms involved in the process so you maintain your position as the smartest person in the room.

 

NLP : Stands for Natural Language Processing, this is the science concerned with understanding what people say and convert it into computer language. Example:

Example:

In his natural way of saying things. A user says: “How much did i spend last month?”

Using NLP, the system will analyze this sentence and turn it into something like select transactions from account where account_number=’123456′ and value<0 and 1/8/2017<transaction_date<1/9/2017

Conversational Interface: This is where conversations between human and a chatbot takes place, usually it’s on a messaging platform like Facebook Messenger, Skype, Telegram, etc. Giving your company or system a conversational interface means allowing people to talk/chat to your system to get answers or perform functions, as opposed to clicking buttons or filling forms.

 

Intent: What the user wants, what’s the purpose behind the sentence the user said. In general, you will need to define the intents the users might have when talking to your bot.

 

If it’s a restaurant bot, intents may include the following:

Order

Location_Inquiry

View_menu

View_Promotion

Reserve

 

Entity: The specific item the intent is about.

I want to order Pizza. Following the example above, Order is the intent and Pizza is the entity.

User Says: The list of anticipated sentences the user might say to convey an intent.

In the example above, all the following sentences (User Says) will convey the intent of view_menu.

Show me the menu

Can I see the menu

What dishes do you offer

What’s on your menu

 

And so on.

The more examples of these sentences you add, the better your bot will understand different variations of the sentences the users say, and convert them into intents that are connected to clear actions, in the case above displaying the menu.

 

Bot Training: The process of feeding your bot with intents, entities and anticipated “user says” sentences in order to make it intelligent enough to converse with your audience.

You’re good now to sit with chatbot experts. If you want to be better than them, go on and create a free account on our bot building platform labiba.ai, you should be able to create great intelligent conversations with ease.