Comprehend Flashcards

(23 cards)

1
Q

Comprehend

TL;DR Comprehend?

A

NLP, output entities in a document

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Comprehend

Is Comprehend real-time?

A

Yes for small jobs. Large jobs need async interface.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Comprehend

What are 4 outputs from a Comprehend job?

A

Entities, language, PII, sentiment

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Comprehend

What 2 ML models does Comprehend use?

A

Pre-trained by AWS, or your own custom model

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Comprehend

How can you trust what Comprehend generated?

A

Every entity has a confidence number letting you know how sure Comprehend is

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Comprehend

What can Comprehend extract from input text?

A

Key phrases, places, people, brands, events, …

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Comprehend

What is sentiment in Comprehend?

A

How positive or negative the text is

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Comprehend

What does Comprehend do with topics?

A

Comprehend can organize text by topic

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Comprehend

How can you ask Comprehend to group files by your own categories?

A

Create training data, give to Comprehend, it creates a custom Classifier

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Comprehend

What is NER in Copmprehend?

A

Named Entity Recognition

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Comprehend

What is NER doing?

A

Find people, places, etc.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Comprehend

How can Comprehend recognize my own (custom) entities?

A

Train Comprehend with your own input data as a Custom Entity Recognizer

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Comprehend

What’s an example of an Entity?

A

“July 31st”, “Date”, 99% confidence

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Comprehend

What common labels are available for Entities?

A

Person, Place, Organization, Date, Quantity

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Comprehend

What’s a problematic Label that Comprehend uses for Entities?

A

“Other”

Doesn’t explain what Comprehend thinks it is, just that it’s noteworthy

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Comprehend

What’s an example of a Key Phrase?

A

“your bank account number 234-957-928364”

It’s all these words together

17
Q

Comprehend

What’s the difference between an Entity and a Key Phrase?

A

Entities are the building blocks, Key Phrases are higher-level constructs

18
Q

Comprehend

Can Comprehend detect PII?

A

Kinda: yes, but it isn’t great, it flags “July 31st” as PII as well as arbitrary numbers.

19
Q

Comprehend

What is “Syntax” in Comprehend?

A

Flags each word as verb, noun, pronoun, adjective, etc.

20
Q

Custom Classifier

What’s a custom classifier?

A

Way to decide what a whole document is about

21
Q

Custom Classifier

Example of custom classifiers?

A

invoice, PR/FAQ, security review

the type of the whole document

22
Q

Custom Classifier

How do you create a custom classifier?

A

Provide lots of Labeled training data in S3, tell Comprehend to build a model from your data

23
Q

Custom Classifier

What’s the architecture for a Comprehend custom classifier?

A

You’re training a new model that you have to deploy with Endpoints, then call the endpoints