Selecting and incorporating built-in skills is a crucial aspect of preparing to take the AI-102 Designing and Implementing a Microsoft Azure AI Solution exam. By understanding these skills and how they apply specifically to manipulating and understanding documents, candidates are better equipped to design, implement, and manage AI solutions in Microsoft Azure.
There are a variety of built-in skills available for working with documents. From text recognition to form readers, these skills are designed to facilitate and simplify the process of creating complex text mining workflows. These built-in skills make it easy to extract vital information from unstructured documents, convert documents and forms into structured data, and enrich this data with knowledge.
Text Recognition and Form Recognisers
Text recognition and form recognisers are built-in skills that help in the automatic extraction of data from documents. Text recognition is designed to execute optical character recognition (OCR) on text, thereby rendering scanned or handwritten documents searchable.
On the other hand, form recognisers are designed to extract data from structured documents with fixed field locations, like invoices or tax forms. They are pre-trained with machine learning models to recognise and extract insights from the entered text.
{
"@odata.type": "#Microsoft.Skills.Vision.OcrSkill",
"context": "/document/normalized_images/*",
"defaultLanguageCode": "en",
"detectOrientation": true,
"inputs": [
{
"name": "image",
"source": "/document/normalized_images/*"
}
],
"outputs": [
{
"name": "text",
"targetName": "ocr_text"
}
]
}
The code sample above, from Microsoft’s official documentation, demonstrates this. It sets the context to normalized images within the document and the default language code to English. The input “image” is applied to detect the text’s orientation, whereupon the output gives the OCRed text.
Language Detection and Text Translation
Language detection and text translation skills come in handy when working with multilingual documents. Language detection identifies the language in which the document is written, while text translation translates the text from one language to another.
A structure (also cited from Microsoft’s official documentation) as shown below simplifies this concept.
{
"@odata.type": "#Microsoft.Skills.Text.TranslationSkill",
"context": "/document",
"inputs": [
{
"name": "text",
"source": "/document/content"
},
{
"name": "toLanguageCode",
"source": "/document/languageCode",
"defaultValues": ["en"]
}
],
"outputs": [
{
"name": "translatedText",
"targetName": "translation"
}
]
}
In the above example, the code receives the text within the document as input, translates it based on the defined toLanguageCode (which is set as English here), and gives the translated text as output.
It is worth considering that using these built-in Azure skills with documents may not automatically return a completed AI solution. They are designed to be building blocks that can help shape the overall solution. With careful understanding and correct manipulation of these built-in skills, handling documents becomes more effortless, making the process more efficient.
For the AI-102 Designing and Implementing a Microsoft Azure AI Solution examination, an understanding of when and how to use these built-in document skills is crucial. Therefore, refining your understanding and expanding your knowledge with the help of resources such as Microsoft’s documentation is a recommended step in the learning journey for the AI-102 exam.
Practice Test
True or False: Built-in skills in Azure AI can be utilized for document search.
- True
- False
Answer: True
Explanation: Built-in skills provide text processing capabilities that can be used, for example, to detect organization names or locations within text, translate text into different languages, or search through the documents.
Which of the following are built-in skills that can be used for documents in Azure AI?
- A. Entity recognition
- B. Language detection
- C. Textual Entailment
- D. Key phrase extraction
Answer: A, B, D
Explanation: The Entity recognition, Language detection, and Key phrase extraction are examples of built-in skills for documents in Azure AI.
True or False: Custom skills can be included in your AI solutions along with built-in skills.
- True
- False
Answer: True
Explanation: In addition to the built-in skills, Azure AI allows developers to create and use custom skills tailored to their specific needs.
If you want to extract key phrases from text, which built-in skill would you use in Azure AI?
- A. Language Detection
- B. Key Phrase Extraction
- C. Merge
- D. Split
Answer: B. Key Phrase Extraction
Explanation: The Key Phrase Extraction skill evaluates unstructured text, and for each record, returns a list of key phrases.
True or False: The Merge skill can be used to combine multiple text values into a single value.
- True
- False
Answer: True
Explanation: The Merge skill in Azure AI merges or concatenates text from a list of documents into a single text value.
Which of the following built-in skills is used to identify the language of text in Azure AI?
- A. Image analysis
- B. Text Translation
- C. Language Detection
- D. Text speech synthesizer
Answer: C. Language Detection
Explanation: The Language Detection skill identifies the language of input content and reports a single language code for every document submitted.
True or False: Built-in skills in Azure AI require coding expertise to be implemented.
- True
- False
Answer: False
Explanation: Built-in skills in Azure AI are pre-built models that do not require any coding to use.
Which built-in skill is used to generate a simplified version of a text in Azure AI?
- A. Language Detection
- B. Text Simplification
- C. Text Translation
- D. Text Normalization
Answer: B. Text Simplification
Explanation: Text Simplification Skill takes text as input and generates a simplified version of said text.
True or False: All built-in skills in Azure AI support multilingual document processing.
- True
- False
Answer: False
Explanation: Not all built-in skills in Azure AI support multilingual document processing. Language support depends on a specific skill.
If you want to link entities in text with more details on the web, which built-in skill would you use in Azure AI?
- A. Web Entity Recognition
- B. Entity Linking
- C. Entity Recognition
- D. Entity Extraction
Answer: B. Entity Linking
Explanation: Entity Linking Skill helps in recognizing and identifying links for entities in the text with more details available on the web.
Interview Questions
What are built-in skills in Microsoft Azure?
Built-in skills in Microsoft Azure are pre-developed functions that accomplish a specific task. They are step in a knowledge mining pipeline, processing from an input to an output.
Can built-in skills for documents in Microsoft Azure be modified?
No, built-in skills for documents in Microsoft Azure cannot be modified. If a skill that accomplishes a specific task is needed but cannot be found within the available built-in skills, a custom skill can be built.
What is an example of a built-in skill in Azure?
An example of a built-in skill in Azure is the OCR skill. This function will identify text within images.
Can built-in skills in Azure be used outside of a skillset?
No, built-in skills can only be used within a skillset in Azure.
How do built-in skills for documents help in Microsoft Azure AI Solutions?
Built-in skills for documents help in refining and structuring the data by processing and transforming the data into information that can be actionable.
What kind of data or content can be processed by built-in skills in Azure?
Built-in skills can process multiple types of content including images, text, and mixed content.
What is the significance of the OCR skill in Azure?
The OCR (Optical Character Recognition) skill in Azure is significant as it allows the extraction of text from image files, which can be further analyzed or stored for future reference.
What are skillsets in Microsoft Azure?
A Skillset in Microsoft Azure is a collection of skills (built-in or custom) that carry out a sequence of information processing tasks.
Can you sequence the operations performed by built-in skills in Azure?
Yes, you can sequence the operations performed by built-in skills by meticulously structuring them within a skillset.
What are custom skills in Azure?
Custom skills in Azure are skills that are built to cater specific needs unmet by the available built-in skills. They are web API endpoints which accept specific input and produce specific output.
Can we combine built-in skills and custom skills in a single skillset in Azure?
Yes, built-in skills and custom skills can be combined in a single skillset to fulfil complex information processing demands.
What is Azure Cognitive Search?
Azure Cognitive Search is an AI-powered cloud search service for mobile and web app development that fine-tunes content based on a user’s preferences and habits.
What is the role of built-in skills in Azure Cognitive Search?
Built-in skills in Azure Cognitive Search helps in defining the steps in an indexing pipeline that extracts, processes, and composes content from unstructured and semi-structured data.
How do you include built-in skills for a document in Azure?
In Azure, you can include built-in skills for a document by defining them in the skillsets definition during the creation of a skillset.
What is the entity recognition skill in Azure and how is it used?
The entity recognition skill in Azure is a predefined skill that identifies and categorizes entities within text. This is widely used for processing customer reviews, analyzing user feedback, etc.
extutorials.com