Amazon Comprehend and Rekognition

Amazon Comprehend

Natural Language Processing (NLP) service.

Key Features

Feature	Description
Entity Recognition	Identify people, places, organizations
Sentiment Analysis	Positive, negative, neutral, mixed
Key Phrases	Extract important phrases
Language Detection	Identify language
PII Detection	Find personal information
Custom Classification	Train custom models
Custom Entities	Train custom NER

Example Usage

import boto3

comprehend = boto3.client("comprehend")

# Sentiment analysis
response = comprehend.detect_sentiment(
    Text="I love this product! It's amazing.",
    LanguageCode="en"
)
# Returns: {"Sentiment": "POSITIVE", "SentimentScore": {...}}

# Entity detection
response = comprehend.detect_entities(
    Text="Amazon was founded by Jeff Bezos in Seattle.",
    LanguageCode="en"
)
# Returns entities: Amazon (ORGANIZATION), Jeff Bezos (PERSON), Seattle (LOCATION)

# PII detection
response = comprehend.detect_pii_entities(
    Text="My email is john@example.com and phone is 555-1234.",
    LanguageCode="en"
)

Comprehend Medical

Healthcare-specific NLP.

• Medical entity extraction
• RxNorm codes
• ICD-10 codes
• Protected health information (PHI)

---

Amazon Rekognition

Computer vision service.

Key Features

Feature	Description
Object Detection	Identify objects in images
Face Detection	Detect and analyze faces
Face Comparison	Compare faces
Celebrity Recognition	Identify celebrities
Text Detection	Extract text from images
Content Moderation	Detect inappropriate content
Custom Labels	Train custom object detection

Example Usage

import boto3

rekognition = boto3.client("rekognition")

# Label detection
response = rekognition.detect_labels(
    Image={"S3Object": {"Bucket": "bucket", "Name": "image.jpg"}},
    MaxLabels=10,
    MinConfidence=80
)

# Face detection
response = rekognition.detect_faces(
    Image={"S3Object": {"Bucket": "bucket", "Name": "photo.jpg"}},
    Attributes=["ALL"]
)

# Content moderation
response = rekognition.detect_moderation_labels(
    Image={"S3Object": {"Bucket": "bucket", "Name": "image.jpg"}}
)

Video Analysis

# Start video analysis
response = rekognition.start_label_detection(
    Video={"S3Object": {"Bucket": "bucket", "Name": "video.mp4"}},
    NotificationChannel={
        "SNSTopicArn": sns_topic_arn,
        "RoleArn": role_arn
    }
)

# Get results
response = rekognition.get_label_detection(JobId=job_id)

Exam Focus Areas

!!! warning "Key Topics" - Comprehend for text analysis and NLP - Rekognition for image/video analysis - Custom models for domain-specific use cases - Content moderation capabilities - Healthcare-specific Comprehend Medical