A user’s intent may be clear but still too broad to answer well. An entity adds the contextual detail that refines an intent, such as a destination, fare type, or product specification. section 4.1.3 section 3.3.1
Traditional conversational AI identifies an intent, or an intent–entity combination, and then selects an action. That action may answer a question, request missing information, or begin a transaction. section 4.1.3
From broad request to useful next step
“I want to book a flight” expresses the Book_Flight intent but supplies no destination, so the assistant collects it. “I need a one-way ticket to Costa Rica” expresses the same intent and provides the Costa Rica entity, allowing the assistant to collect departure details. Similarly, “I’d like to upgrade my seat to first class” uses first class to refine the Flight_Upgrade intent and initiate the upgrade process. section 4.1.3
Table 4.1 Example utterances may be handled differently based on the presence or absence of entities.
|
Utterance
|
Intent
|
Entity
|
Possible next step
|
|---|---|---|---|
|
“How many bags can I check?”
|
Bag_Allowance
|
Display bag check policy
|
|
|
“I want to book a flight”
|
Book_Flight
|
Collect destination
|
|
|
“I need a one-way ticket to Costa Rica”
|
Book_Flight
|
Costa Rica
|
Collect departure details
|
|
“I’d like to upgrade my seat to first class”
|
Flight_Upgrade
|
first class
|
Initiate upgrade process
|
Compare utterances where the same intent leads to different handling because an entity is absent or present. section 4.1.3
Use entities when additional context makes an answer more specific or an interaction more actionable. Their value should follow user demand and business goals rather than an attempt to cover every possible topic. section 4.1.3 section 3.3.1
