An Overview of Microsoft Azure NLP Solutions
In this post, I will briefly discuss AI and NLP timelines with an overview of Azure AI. I with then provide a guideline of how to map Azure AI solutions with NLP techniques.
The target audience for this post are developers, data scientists and machine learning engineers with different levels of NLP knowledge who are seeking an easy guide to try Azure AI/NLP solutions.
What is NLP?
Natural Language Processing (NLP) is a field of artificial intelligence, machine learning and computational linguistics. Its primary goal is the interactions between computers and human (natural) languages including text and speech. In particular, NLP provides different techniques of how a computer program can understand, analyse, and potentially generate large volume of human language data.
NLP applications include natural language understanding, machine translation, semantics , the syntactic passing, natural language emulation, dialectal systems such as speech recognition, question answering and a broad range of text analytics techniques such as topic modelling, classification, summarization, sentence/document similarities etc.
Brief History of AI and NLP
The term AI was coined by American computer scientist John McCarthy at the 1956 Dartmouth Conference, widely considered the birthplace of the discipline. Known as the father of AI, McCarthy created the Lisp computer language in 1958, which became the standard AI programming language and continues to be used today.
The breakthrough in AI -and NLP in particular- happened when Alan Turing, an English mathematician published his famous article Computing Machinery and Intelligence and proposed what is now called the Turing test.
Turing test is used to determine whether or not computer(machine) can think intelligently like human?. This criterion depends on the ability of a computer program to impersonate a human in a real-time written conversation with a human judge, sufficiently well that the judge is unable to distinguish reliably — on the basis of the conversational content alone — between the program and a real human.
Below the AI-timeline is captured at a glance with the breakthroughs of Machine learning, Deep learning and NLP:
Microsoft Azure AI
The Microsoft AI platform provides a suite of powerful tools to allow developers to easily and quickly infuse AI into their applications and scenarios, enabling new, intelligent experiences for their users. Azure AI is divided into three main pillars:
Below is the official repositories for some of the services published by Microsoft:
1) Azure cognitive services:
These solutions are there APIs, SDKs, and services available to help developers build intelligent applications without having direct AI or data science skills or knowledge. Azure Cognitive Services enable developers to easily add cognitive features into their applications. The goal of Azure Cognitive Services is to help developers create applications that can see, hear, speak, understand, and even begin to reason. The catalogue of services within Azure Cognitive Services can be categorized into five main pillars — Vision, Speech, Language, Web Search, and Decision. Below are the links to the official repositories on Microsoft. Please note that for accessing the repo, the users need to request access from Microsoft Open Source github repository.
Microsoft docs resource API access
Project page for Cognitive Services offerings and demos
Free trial access key link
Cognitive Services Java SDK samples
Cognitive Services Node.js SDK samples
Cognitive Services Python SDK samples
Cognitive Services Speech SDK
Cognitive-services-REST-api-samples
Azure Cognitive TTS Samples
Computer Vision Recipes
Language Understanding (LUIS)2) Knowledge Mining solutions:
These solutions are designed for advanced knowledge mining tasks, such as Name Entity Recognition, Phrase Extraction, Custom labelling and Custom skills, where can enrich the IE process. Below are the links to the official repositories on Microsoft. Please note that for accessing the repo, the users need to request access from Microsoft Open Source github repository.
Azure Cognitive Search Accelerator
Form Recognizer Recipes3) Deep Learning NLP solutions:
Microsoft research has published a repository contains examples and best practices for building NLP systems, provided as Jupyter notebooks and utility functions. The focus of the repository is on state-of-the-art methods and common scenarios that are popular among researchers and practitioners working on problems involving text and language. Please note that for accessing the repo, the users need to request access from Microsoft Open Source github repository.
Microsoft NLP Recipes
Mapping NLP applications to Azure AI
As an NLP specialist for five-year and a Microsoft Cloud Solution Architect for two-year, I have come up with a mapping of NLP applications to Azure AI solutions. This mapping can be used as a guideline for customers and partners who has chosen Azure cloud for their NLP scenarios.
At the end, let me conclude this post by looking at Microsoft AI Platform at a glance:
The content of this post is also available through the public github repo with the links to the Microsoft official repositories, there is a webinar recorded for this talk, available to the public demand.
#azure #azureml #datascience #nlp #machinelearning #deeplearning #microsoft
