Page 379 - Ai_V3.0_c11_flipbook
P. 379
For instance, assume you’re developing a chatbot for a hotel chain. A common user’s intent would be to ask about room
availability. To ensure the chatbot can effectively handle this intent, you might ask a front desk agent who frequently
answers these questions in person or over the phone. Your objective would be to compile a comprehensive list of the
various ways guests might inquire about room availability. The table below includes different examples of potential user
inputs related to this intent.
Intent Possible user inputs
Do you have any rooms available?
Can I book a room for this weekend?
Do you have any available rooms for next week?
Check Room availability Is there a room available for two people?
Are there extra charges for kids?
Kids of what age group can stay in the same room?
I want to book a room in your hotel. Is it available this weekend?
Entity
An entity is a noun that represents a person, place, or thing. Once you’ve compiled a list of the intents you want your
chatbot to satisfy, you proceed further. For instance, if a user asks, “Do you have any rooms available in Delhi?”, then
checking room availability is the intent and Delhi is the entity. A chatbot requires a complete set of entities in order to
be useful. The table below contains instances of entities that correspond to the intent and probable user inputs in the
previous hotel example:
Intent Possible user inputs Entities
Do you have any rooms available in Location, Delhi
Delhi?
Check Room Availability Is there any room free for tonight? Date
Is there a room available for two Number of guests
people?
Can I book a room in Mumbai for Location and Date
next week?
Book a Room Can I reserve a room for tomorrow? Date
I want to book a family room for
this weekend. Room Type and Date
Leveraging Linguistics and Computer Science 377

