Build a bot with Azure Bot Service - GOL Clinics Recap
Published Aug 15 2022 01:32 AM 1,104 Views
Microsoft

Game of Learners Clinics for ML (Machine learning) and AI (Artificial intelligence) is a 5-week Skilling initiative for students to level up in Artificial Intelligence on Azure. Find out more on previous sessions at http://aka.ms/golaiml-home  

Natural Language Processing (NLP) involves analyzing text documents or phrases to gain insights into the content of the text.  

In this session you explore natural language processing and perform text analysis using Azure text analytics service. Additionally, you will build a bot with Language Service and Azure Bot Service. 

 

Azure Resources for Language Services: 

bethanyjep_0-1658746769430.png

Using Microsoft Azure, you can perform several actions to your text including language detection, sentiment analysis, key phrase extraction, and entity recognition. A service that will come in handy to help you perform this is the Language Cognitive Service. 

 

Language Cognitive Service  

The Language service is a part of the Azure Cognitive Services offerings that can perform advanced natural language processing over raw text. As you analyze text/documents using the Language Cognitive Service, you can: 

  • Determine the language of a document whether it is in English or Swahili 
  • Perform sentiment analysis to determine whether the text is positive, negative or neutral. Its usefulness can be in determining user feelings towards a brand or a specific topic. 
  • Extract and categorize phrases from text that may point to specific topics. 
  • Identify and categorize entities in text. Entities include people, places, organizations and everyday items like dates or quantities. 

Azure Bot Service 

Imagine you subscribe to a service and the only way you can contact them is through their FAQ. What would you do if you needed to cancel your subscription? This is where Conversational AI comes in. The Azure Bot Service provides a framework for developing, managing and publishing bots on Azure. 

 

The first step is creating a knowledge base. You can use Language Studio’s custom question answering feature to create, train, publish and manage knowledge bases. 

bethanyjep_1-1658746769434.png

 

Applications of NLP  

How can you apply NLP in real world applications? 

  • Feed Analyzer: a social media feed analyzer to detect sentiments around a brand or a political campaign 
  • Document Search: a document search application to extract key phrases and help summarize the main subject matter of the document in a catalog. 
  • Text Extraction: a tool to extract information or company names from documents or other text for identification purposes. 

Reference and Resources: 

Follow along and build your Bot with Azure Bot Service at: https://aka.ms/gol-nlp  

 

Co-Authors
Version history
Last update:
‎Aug 15 2022 01:32 AM
Updated by: