Natural language processing solutions 25-30%

Question 1

Read OCR

Answer 1

Extract print and handwritten text including words, locations, and detected languages

Question 2

Layout

Answer 2

Extract text and document layout elements like tables, selection marks, titles, section headings, and more

Question 3

General

Answer 3

Extract key-value pairs in addition to text and document structure information

Question 4

Health insurance card

Answer 4

Prebuilt: automate healthcare processes by extracting insurer , member, prescription, group number, and other key information from US health insurance cards

Question 5

w-2

Answer 5

Prebuilt: process w2 forms to extract employee, employer, wage, and other information

Question 6

invoice

Answer 6

Prebuilt: automate invoices

Question 7

Receipt

Answer 7

Prebuilt: extract receipt data from receipts

Question 8

Identity document

Answer 8

Prebuilt: extract identity fields from US driver licenses and international passports

Question 9

Business card

Answer 9

Prebuilt: scan business cards to extract key fields and data into your applications

Question 10

Form Recognizer prebuilt models

Answer 10

Health insurance car, W-2, invoice, receipt, identity document, business card

Question 11

Form recognizer custom models

Answer 11

Custom extraction models, custom classification models, composed models

Question 12

Translations services

Answer 12

Language detection, one to many translation, transliteration

Question 13

Preconfigured features of a language understanding app

Answer 13

Summarization, named entity recognition, PII detection, key phrase extraction, sentiment analysis, language detection

Question 14

Learned features of a language understanding app

Answer 14

conversational language understanding, custom-named entity recognition, custom text classification, question answering,

Question 15

What three parameters will all conversational language understanding requests have

Answer 15

kind (type), parameters, analysis input

Question 16

How to fix substitution errors?

Answer 16

add custom product and people names to the training data

Question 17

What type of error do you get with overlapping speakers?

Answer 17

deletion errors

Question 18

What type of error do people talking in the background give you?

Answer 18

insertion errors

Question 19

What is text independent verification?

Answer 19

The app will identify voices of people without a special verification phrase

Question 20

What is text dependent verification?

Answer 20

The app will identify voices with a passphrase

Question 21

What are the speech to text models?

Answer 21

base, custom, adapted

Question 22

When should you use a custom acoustic speech to text model?

Answer 22

Noisy areas

Question 23

When should you use a custom language speech to text model?

Answer 23

topics like biology, physics, radiology, product names, and custom acronyms

Question 24

What is the BLEU score?

Answer 24

An algorithm for evaluating the precision or accuracy of text that has been machine translated from one language to another.

Question 25

What are the bounds of the BLEU score?

Answer 25

zero to one hundred

Question 26

What range of scores indicates a high quality BLEU score?

Answer 26

40-60

Question 27

What are not allowed in synonyms?

Answer 27

special characters like #