Azure AI Language

Azure AI Language Cheat Sheet

Azure AI Language is a suite of natural language processing (NLP) services under Azure Cognitive Services. It helps developers build applications to understand, analyze, and generate human language. The services support language understanding, entity recognition, sentiment analysis, translation, and text analytics.

It provides SDKs and REST APIs that integrate easily with applications across different languages and platforms.

Features	Functionality
Named Entity Recognition (NER)	Detects and categorizes entities such as people, organizations, and locations within text.
Personal and Health Data Detection (PII Detection)	Identifies personally identifiable information in text and conversational data to protect sensitive personal and health details.
Language Detection	Automatically determines the language and dialect of the input text from a broad set of supported languages.
Sentiment Analysis and Opinion Mining	Analyzes text to assess sentiment polarity (positive, unfavorable, neutral, mixed) and links sentiments to specific text elements, aiding brand or topic perception analysis.
Summarization	Generates concise summaries of text and conversations using two approaches: Extractive: Selects key sentences from the original content. Abstractive: Produces new, coherent sentences summarizing the content. Supports conversation segmentation and summarization of customer interactions.
Key phrase extraction	A preconfigured feature that evaluates and returns the main concepts in unstructured text, and returns them as a list.
Entity Linking	Resolves ambiguities of entities in text by linking them to external knowledge bases like Wikipedia.
Text Analytics for Health	Extracts and labels medical and health-related information from unstructured clinical text.
Custom Text Classification	Enables building tailored AI models to categorize unstructured documents into user-defined classes.
Custom Named Entity Recognition (Custom NER)	Allows creation of models that detect custom entity types based on user-provided training data.
Conversational Language Understanding (CLU)	Builds custom models to interpret user intent and extract relevant information from conversational utterances.
Orchestration Workflow	Integrates CLU, question answering, and LUIS apps into seamless workflows for complex conversational scenarios.
Question Answering	It answers user questions precisely and is ideal for chatbots, social media, and voice-enabled apps.

Key Benefits:

Comprehensive NLP Capabilities – Supports a wide range of language tasks, from simple language detection to complex conversational understanding.
Customizable Models – Enables building and training custom entity and intent models tailored to specific business domains.
Multi-Language Support – Supports over 100 languages and dialects for analysis and translation.
Scalable and Reliable – Built on Azure’s global infrastructure with high availability and low latency.
Seamless Integration – Easy integration with Azure services and other applications through REST APIs and SDKs.
Security and Compliance – Offers enterprise-grade security and meets major compliance certifications, ensuring safe processing of sensitive data.
Cost-Effective – Flexible pay-as-you-go pricing with free tier options for trial and development.

Continuous Updates – Microsoft’s AI research powered regular improvements and feature additions.
Real-Time and Batch Processing – Supports real-time API calls and batch text processing for large datasets.

Security:

Data Privacy: Customer data is not used for training Microsoft models unless explicitly opted in.
Compliance: Azure AI Language complies with GDPR, HIPAA, FedRAMP, and other global standards.
Data Encryption: Data is encrypted in transit and at rest using industry-standard protocols.
Role-Based Access Control (RBAC): Access to AI Language resources can be managed via Azure Active Directory roles and policies.
Private Endpoint Support: Use Azure Private Link to access the service over private IP addresses, avoiding exposure to the public internet.
Network Security: Integration with Virtual Network (VNet) service endpoints and firewall rules for controlled access.
Audit Logging: Integration with Azure Monitor and Azure Security Center for auditing and threat detection.

Pricing:

The pay-as-you-go model is based on the number of text records or transactions processed.
Different pricing tiers available for:
- Text Analytics (per 1,000 text records)
- Conversational Language Understanding (per prediction request)
- Translator (per character count)
- Custom Models: Additional charges for training and deployment of custom entity models.
Free tier available for limited usage, typically includes:
- Up to 5,000 transactions per month for Text Analytics
- Limited LUIS calls and Translator characters
- Enterprise agreements and committed use discounts are available for extensive volume usage.
- Pricing varies by region.

For detailed pricing, refer to: https://azure.microsoft.com/en-us/pricing/details/cognitive-services/language-service/

References:

Written by: Ace Kenneth Batacandulo

Ace is AWS Certified, AWS Community Builder, and Junior Cloud Consultant at Tutorials Dojo Pte. Ltd. He is also the Co-Lead Organizer of K8SUG Philippines and a member of the Content Committee for Google Developer Groups Cloud Manila. Ace actively contributes to the tech community through his volunteer work with AWS User Group PH, GDG Cloud Manila, K8SUG Philippines, and Devcon PH. He is deeply passionate about technology and is dedicated to exploring and advancing his expertise in the field.