Deep Learning
Document Understanding
Description
Document Understanding operation is designed to extract the answer of questions asked from the provided text. Text can be data from pdf or image or any other file format. We can ask multiple question at a time for provide text. Text can be processed text and as well as raw text.
Number of Parameters : 4
Parameter : Processing Text
Provide the name of key which contains the text inside a square bracket.
Example:
['processed_text']
We can provide nested JSON key
Example:
['data_response']['processed_text']
This parameter cannot contain more than one key name inside square bracket
Example:
['processed','text']
Parameter : Question?
Provide here the list or array of question or questions whose answers are required to be extracted from the text file.
Example:
What is the name of State?
Which is the capital of State?
What is the language of State?
What is population of State?
Parameter: Attribute?
Enter the attributes' name against each question that will contain the answer to the question asked.
Example:
If the question asked is: What is the name of State?
Enter the attribute as:
State
Parameter: Understanding Key?
Provide the key name that will store the answers of all the given questions.
Example:
Answers
Image Understanding
Description
This operations helps to extract important information from any Image based data by asking questions.
Number of Parameters : 4
Parameter : Processing Image
Provide the key name having the base64 code of image from source.
Example:
['processed_text']
Parameter : Questions
Provide questions of information that is required to be extracted from Image.
Example:
What is the name of State?
Parameter: Attribute
Provide Attribute name against each question to save the answers of asked question.
Example:
If the question asked is: What is the name of State?
Enter the attribute as:
State
Parameter: Understanding Key
Provide the key name that will store the answers of all the asked questions.
Example:
Answers