Conversation
Linguistically...
A Motherload of Data
Probabilistic Modeling
100

The maxim requiring speakers to be relevant.

What is the Maxim of Relation?

100

This type of speaker will have a different dominant language than the one spoken at home, and may not be fluent in the home language.

What is a heritage speaker?

100

This type of data is not standardized.

What is unstructured data?

100

Probability (chance) of going from one letter to the next.

What is transition probability?

200

This is the name of the first chatbot.

What is ELIZA?

200

This language has the most second language speakers in the world.

What is English?

200

Voice transcriptions and social media fall under this classification of data type.

What is unstructured data?

200

Probability of one letter being mistyped for another.

What is confusion probability?

300

When the hearer of a conversation acknowledges that they understand the speaker, they do this.

What is grounding (the conversation)?

300

A group of words that function together as a unit.

What is a constituent?

300

An intelligent virtual assistant (e.g. Siri or
Alexa).

What is an example of a Dialogue Agent / Assistant?

300

A sequence of n words.

What is an n-gram?

400

The statement of "Well, it depends on the weather and where the cow came from" to the question of "How do you like your steak?" violates this maxim.

What is the Maxim of Quantity?

400

This type of error correction corrects errors or suggests corrections without taking the surrounding context into account.

What is isolated-word error correction?

400

Blogs are this.

What is semi-structured data?

400

Models that assign probabilities to upcoming words, or sequences of words.


What are language models (LMs)?

M
e
n
u