Getting Started with AI using IBM Watson

Week1

Introduction

 
In Module 1, you'll understand what Watson is and how Watson AI works. You
will understand the many ways Watson AI is helping professionals and
businesses to re-imagine their workflows, and leveraged AI to deliver value to
their clients.

 
In Module 2, you will learn about some of the Watson AI services offered on
the IBM Cloud. You will understand how organizations can use Watson AI
services, and the types of situations in which each service applies. You will
specifically learn about services like Watson Assistant, Watson Discovery and
Watson services related to knowledge, speech, and language.

 
In Module 3, you will learn about a few more Watson AI services including
services related to data, 
vision, empathy, lifecycle, and compliance. We'll also be able to understand
what kinds of applications these services can be used for.

in Module 4, you will learn about commonly used cases for AI and look at
some case studies involving Watson AI. You could also experience and
demonstrate AI in action yourself using Watson.

Natural language and conversation APIs allow me to understand the many

project related to questions which employees can ask via their tablet or
computer. All they must do is ask. Using my retrieving rank API, I can find
the answers by reading through tens of thousands of project related
documents in seconds. And with Watson Explorer, I can apply expertise that I
have learned from thousands of employees. Working together with
Woodside, I have helped reduce time spent tracking down expert knowledge
by 75 percent.
Watson isn't just the promise of AI, is the real deal deep learning business
AI, that knows your industry, 
protects your insights and works with tools that you already use.
How Watson Works

Watson is an artificial intelligence system. Artificial Intelligence is what gives

machines the power to learn, adapt to new inputs, and make better
decisions. Machine Learning, a subset of AI, uses computer algorithms to
analyze data and make intelligent decisions based on what it's learned. It's
how streaming sites recommend new music by comparing what one
listener likes to others with similar tastes. Watson uses a sophisticated
machine learning technique called Deep Learning. Deep Learning layers
algorithms to create an artificial neural network. It can continuously learn on
the job, determining whether decisions were correct, constantly improving the
quality and accuracy of results. This is what enables it to learn
from unstructured data such as photos, videos, and audio files. Meaning
Watson can tell an insurance company exactly how a car's been damaged in a
crash, by discerning the make and model before 
comparing images of an undamaged car to the damaged one, even getting as
detailed as a broken exhaust. Deep Learning is also what enables Watson's
Natural Language Understanding capabilities, 
allowing it to learn by deconstructing sentences, then analyzing and
identifying the concepts and relations of those sentences. Once it understands
how everything fits together, it can work out the context and intent of what's
being conveyed. But your AI also needs to understand the specific language
and terminology of your industry. With traditional AI models, this would require
lots of additional data and computing power, taking weeks or months. With
Watson however, it can be done quickly and simply, 
thanks to a technique called Transfer Learning.

Transfer Learning is learning how to learn. It’s what enables Watson to learn
more from less. So, it doesn't need to be trained from scratch. It can be fed
prior knowledge to speed things up. Let's take the case of a home insurance
company, to demonstrate how Watson's Transfer Learning is a three-layered
AI model. The bottom layer is made up of out of the box general
knowledge, like Wikipedia for AI. This layer tells Watson what a house is, or a
tornado. The middle layer is pre-packaged with knowledge tailored to specific
industries. So, for insurance, it's how Watson understands terms like coverage
or beneficiaries. 
The top layer is where the customer specific learning takes
place, personalizing Watson to the company’s unique business needs. Fed
with their data and know-how, the model can understand a company's specific
risk attributes, and behavioral policy pricing models. Transfer Learning is a key
part of how Watson is already accelerating claims processing by 25
percent, dramatically cutting operating costs. 
Transfer Learning makes Watson the best AI for your enterprise. It allows it to
learn faster from smaller sets of data, while also protecting your data and
insights. Watson is designed to ensure data only flows from the bottom to the
top. So, you can make the most of the shared knowledge in the bottom two
layers, while retaining ownership of your top layer. Most companies AI
systems use your data to build their general machine learning
models, allowing all their users to benefit from that collective knowledge. 
Watson's three-layer model ensures your data, remains your data.

Lesson Summary

In this lesson, you have learned:

 Examples of real-world applications of Watson in Energy, Natural Gas

Production, Farming, and Insurance.
 Deep Learning enables Watson to learn from unstructured data and
constantly improve the quality and accuracy of results by learning on the
job.
 Transfer Learning enables Watson to learn faster from smaller sets of
data while also protecting customer's data and insights.
 https://www.ibm.com/watson/ai-stories/
 
IBM's Watson Assistant, a virtual assistant powered by underlying natural
language models, 
understands your users and provides training recommendations as you build
and run your assistant. 

Royal Bank of Scotland's home buying and ownership business has employed

CORA and MARGE, both powered by Watson to deliver better customer
service and to empower the context center agents.

When developing CORA, we analyzed data to determine the most commonly

asked questions. I don't recognize this transaction? How do I pay my bills
online? I want to change my address. CORA's trained with over 200 customer
intents and has over a thousand responses to those intents. The simple
questions. I'd like to check my balance? CORA responds with the answer
straight away. For more challenging questions, we offer guidance to the
customer so that they can help themselves. And for complicated
questions: How do I make my first mortgage payment? CORA seamlessly
hands off the customer to an agent. When customers get to me, I have that
information already. I think our AI journey has been fantastic, and it's going to
allow us to create new products and new benefits for our customers for years
to come. Watson Assistant automatically directs requests down the optimal
path for solving the problem. It knows when to provide a direct answer to a
common question or reference more generalized search results for something
more complex. It also helps agents find answers to complex questions
faster. Watson Assistant is powered by underlying natural language models
that understand 
your users and provide training recommendations as you build and run your
assistant.

The foundation of Watson Assistant addresses four key principles. 

 First, it comes pre-trained with industry-relevant content. 
 The second is that, it remembers and understands user input. 
For example, your customer's having a conversation with Watson and Watson
recognize a new intent, Watson needs to answer that question and
remember this conversation just as you would expect a human to remember. 
 Third, it recommends. Watson assesses what is happening in these
conversations and recommends places that may be strengthened by
having other intents. When you provide a few examples of questions, 
intents, and clarifying terms, entities. Watson Assistant can build a machine
learning model to understand similar natural language requests from your
users. 
 The fourth principle is about protecting your data. It ensures your data is
used to train Watson for you and not through your competitors. With
Watson, it is also possible to integrate with customer care platforms, 
making it the best AI for personalizing customer experience.

 IBM Watson Assistant is helping businesses deliver personalized

customer experiences and supporting their contact center agents with the
knowledge and expertise they need to give outstanding customer service.
 IBM Watson Regulatory Compliance has made compliance with rapidly
evolving regulatory requirements into an accurate, timely, trustworthy, and
efficient process.
 AI professionals are building, training, and deploying their custom
models using the Cognitive APIs and end-to-end collaborative environment
provided by IBM.
Most systems will analyze that structured data at a summary level and end up
missing important unstructured information such as the details available in the
free text of an incident report. Watson finds that, exposes it, and organizes the
information to help the safety analyst find workplace safety
improvements. Other companies have put lots into work analyzing
transactions for tax law compliance. 
With the tax law changes, legal teams need to ensure that they are adhering
to all compliances. Watson assists professionals to process mounds of
information, and augment their professional views and 
their professional experience to help them get compliance right and speed it
along. Watson also assists professionals and compliance-oriented business
processes in understanding contracts. Insufficient contracting can cause firms
up to 40 percent of the deals value. Watson has been taught document
structure and contract terminology so that Watson can understand these
contracts and statements of work. Financial institutions are spending more
time and money than ever on regulatory compliance. 
Regulations are changing and increasing rapidly. Regulations, obligations,
policies, and controls are managed in disparate systems that have little to no
relationship to one another, and it's nearly impossible to confirm regulatory
and policy compliance coverage and operating controls suitability to
obligations. IBM Watson Regulatory Compliance is a standalone module of
open pages with Watson the industry's only AI driven GRC solution and
supported by the expertise of Promontory Financial Group. IBM Watson
Regulatory Compliance turns the manual time-consuming and expensive
process of complying with regulatory requirements into a streamlined efficient
process driven by consistent methodology. It combines software, content, and
expertise to provide a holistic, accurate, and timely view of the obligation life-
cycle that enables organizations to achieve trustworthy, efficient, and
sustainable compliance.

At IBM, we have created an end-to-end collaborative environment

that provides a suite of capabilities to help AI professionals build, train,
and deploy their custom models with their data alongside Watson. 
In its simplest form an API (Application program Interface) is a set of
routines, protocols, and tools that help applications work together. Cognitive
APIs do similar work, but they also serve as a messenger between data
and the cognitive computing power of Watson combining these in a coding
platform on the cloud. Cognitive APIs are based on an input output
system. Your input is data that the API feeds into Watson and the output is the
delivery of the overall outcome based on that data. IBM has a huge list of
cognitive APIs that our clients used to make smarter business decisions. Most
cognitive solutions use several APIs together to build apps. Together these
APIs analyze massive amounts of data to reveal what's trending and why. And
since various cognitive APIs can be swapped out to change functionality
and outcomes, they are being used to power many cognitive applications in
almost every industry across the globe. Companies can start with an initial
project and then expand from there and as they expand, they build their
roadmap to AI.

Lesson Summary

In this lesson, you have learned:

 IBM Watson Assistant is helping businesses deliver personalized

customer experiences and supporting their contact center agents with the
knowledge and expertise they need to give outstanding customer service.
 IBM Watson Regulatory Compliance has made compliance with rapidly
evolving regulatory requirements into an accurate, timely, trustworthy, and
efficient process.
 AI professionals are building, training, and deploying their custom
models using the Cognitive APIs and end-to-end collaborative environment
provided by IBM.
 https://www.ibm.com/watson/ai-professionals/

Enterprises do not have the massive data that you see in the consumer
application, they need to be able to learn more from less. Traditional AI
models require massive data and computing power. Watson uses a technique
called transfer learning, that enables it to learn more from less. So, it doesn't
need to be trained from scratch. It can be fed prior knowledge to speed things
up. Watson’s three-layer transfer learning model ensures your data remains
your data. The customer’s specific learning that personalizes Watson to the
company's unique business needs takes place in the top layer. Since Watson
ensures that data only flows from the bottom to the top, it allows the customer
to retain complete ownership of their top layer.

If I can engage an AI in that process of extracting from those articles the things
that are important to me in the context of the problem I'm trying to solve, now
I've gotten an advantage. If I can use AI to help me see through my own
biases or to perspectives that I may not have considered before, now I'm able
to make better decisions because I can now consider those alternatives within
that decision process.
IBM Cloud has been built specifically to maintain the protection and security of
data, to maintain isolation between different institutions that may be using our
cloud or using our technology in the cloud, and we carry that all the way
through to everything that we do. So if I'm summing that back up, it sounds
like the three differences between consumer and business AI really come back
to like embedding into the workflows, putting AI at the fingertips of the person
or the employee when they need it, where they need it so they can get the
vast amount of information faster. It's learning from less data so that you're
able to get started faster and work on this training models, and then finally, the
protection of that data so that it stays with you and doesn't go anywhere else.

Step number one is prove, step number two is adopt, and step number three is
scale.

Lesson Summary

Watson is helping teams and businesses

 Reimagine their workflows, using its capabilities to process massive

amounts of data faster, to remember, understand, and make
recommendations.
 Learn more from less data, using a technique called Transfer Learning.
 Protect their data and insights, using a 3-layered learning model.
It is important to recognize how Consumer AI is different from Business AI

 Consumer AI is all around us, on our smartphones, in our homes and

devices, and impacting virtually all aspects of our everyday lives.

 Business AI impacts how we engage with our clients, the way we

perform our responsibilities at work, and our business processes.
The 3 steps of that can help customers get started on their AI journey

 The use case through which they can prove that AI can benefit their
organization.
 The governance frameworks, methodologies, best practices, and
anything else needed to adopt the proven use case across the
organization.

 Mechanisms for continuous improvement, continuous adoption, and

continuous scale.
 https://www.ibm.com/blogs/watson/2018/10/what-did-your-business-
learn-last-night-3-ways-watson-ai-enables-employees-to-work-at-the-
speed-of-data-and-do-more-interesting-work/
 https://www.ibm.com/blogs/think/2017/12/accelerating-ai/

Week 2

Watson is AI from IBM. Designed to make your business smarter and every
employee your best employee. Watson has evolved into a range of advanced
APIs, specialized tooling and Software as a Service applications, built for the
enterprise. This means that Watson is built for complex use cases and
designed to integrate with platforms that professionals use in their daily work.
Ensuring seamless access to the information you need to make the right
decisions.

The Watson AI portfolio of services, tools, and APIs includes the following.

Watson Assistant helps you build chatbots and virtual assistants for a variety
of channels including mobile devices, messaging platforms, and even robots.

Watson Discovery unlocks hidden value in data to find answers, monitor

trends and service patterns with the world's most advanced cloud native
insight engine.

Watson Studio lets you build train, deploy, and manage AI models, and
prepare and analyze data in a single integrated environment

Watson Machine Learning enables you to use your own data to create, train,
and deploy machine learning and deep learning models.

Watson Visual Recognition allows you to quickly and accurately tag, classify,
and train visual content using machine learning.

Watson Knowledge Catalog drives collaboration and transforms data and AI

into a trusted enterprise asset through dynamic data policies and enforcement.

Watson Knowledge Studio provides you with a way to teach Watson the
language of your domain, with custom models that identify entities and
relationships unique to your industry.

Watson Open Scale provides insights into AI health, recommends next steps
to improve outcomes, and orchestrates tasks to remediate issues around
performance, accuracy, and fairness.
Watson Compare and Comply streamlines contract workflows to save time
and improve accuracy and simplify contract governance.

Watson portfolio also includes Watson Natural Language Understanding.

Watson Natural Language Classifier. Watson Text to Speech. Watson Speech

to Text. Watson Language Translator. Watson Personality Insights. Watson
Tone Analyzer.

Watson AI portfolio enables you to accelerate research, and discovery, enrich

your customer interactions, anticipate and preempt disruptions, recommend
with confidence, scale expertise and learning, and detect liabilities and
mitigate risk.

Watson is the AI solution designed for the realities and constraints of

proprietary enterprise data. Watson was initially available only on IBM cloud,
but is now portable across any cloud powered business. Organizations can
now run IBM Watson services including Watson Assistant, and Watson Open
Scale on any cloud public or private. This prevents customers being locked
into a single vendor and enables them to start deploying AI wherever their
data resides.

I mentioned Watson Discovery. I mentioned the Watson Assistant and the

Watson Voice Agent, but there Is also, for example, the Knowledge Studio.
Now of course, Watson's very extensible, that's really the whole reason it's so
useful. Not only is there the powerful Natural Language Understanding
service, but you can also take that Natural Language Understanding service
and modify it to your own needs. You can find custom entities. You can find
customer relationships. You can do custom correlation detection or co-mention
detection. All these things are possible with the Watson Knowledge Studio.
Apart from that, there is also the Watson Studio, not to be confused with
Watson Knowledge Studio. The Watson Studio used to be called the Data
Science Experience, where you can have the deep learning as a service
environment where you are as a programmer using Jupyter Notebooks and
the visual neural network model in order to create machine learning
applications directly on the Watson environment. So, there are lots of different
services IBM provides, and within Watson, within the fields of vision, within the
fields of audio, within the fields of natural language, within the fields of
customizations.

Lesson Summary

In this lesson, you have learned:

  IBM Watson provides a portfolio services, tools, and APIs to enable
businesses to use and develop AI powered applications.
 The Watson AI portfolio enables you to perform a wide range of tasks,
including accelerate research and discovery, enrich your customer
interactions, scale expertise and learning, and detect liabilities and
mitigate risk.
 Watson is available on any cloud empowered business.
 https://www.ibm.com/watson

The roots of Watson where it's originally evolved from, Watson was originally
a Jeopardy playing machine essentially. That's really what it was
specifically built and targeted to do. And because of that, it has amazing
information retrieval and understanding systems.

The Watson Discovery Service. Now, this is we call the retrieving rank

service. The Watson Discovery Service now enables you to actually feed in
entire corpora of documents, and then from there be able to analyze natural
language within those documents be able to figure out what concepts they talk
about, query them with natural language, find passages from within those
documents, and combining this with chatbots creates a very powerful
experience. Because chatbots by themselves, yes, they're great, and Watson
can do things like for example dynamically understanding entities within
messages, which is another unique feature. But now being able to understand
these depths of contents of the Watson Discovery Service and time that it
would Watson Assistant is something that really is a unique offering and
enables you to take your chatbots to the next level of usefulness.

--------------------------------------------------------------------------------------------------------
Watson Assistant:

Watson Assistant is not another chatbot. Most chatbots try to mimic human
interactions. Frustrating you when a misunderstanding happens. Watson
Assistant gets it. It knows when to search for an answer from a knowledge
base, when to ask for clarity, and when to direct you to a human.
Watson Assistant combines machine learning, natural language
understanding, and integrated dialogue tools to create conversation flows
between your apps and your users. You can use Watson Assistant to
create natural language interfaces to your applications. With it, you can build
artificially intelligent agents that communicate with users through natural
language, making your applications accessible and is easy to use as holding a
conversation. Use Watson Assistant to quickly build and deploy chatbots
and virtual agents across a variety of channels including mobile
devices, messaging platforms, and even robots. Watson Assistant
understands the intent 
of what your users are saying. It can identify entities that users mention and
manage the flow of the conversation. You don't have to be a programmer to
use Watson Assistant. 
The tooling is written with business users in mind and no coding is required to
set up a bot. 
Integrated online help and examples can get your bots running within
minutes. Watson Assistant requires very little data to train an intent. With as
few as five examples, you can have the Watson Assistant understand an
individual intent. Browse from a catalog of already configured customer
service and industry content packs to save time and start faster.

 
Watson comes with this package that lets it respond to basic dialogue. There
are also content packs for general customer service, retail banking, energy
and utilities, insurance, and telecommunications. These content packs can let
you focus on training your assistant on the specifics of your business rather
than having to worry about the general industry knowledge,

Watson Assistant Search Skill offers a unique ability to integrate with AI

powered search. 
It does this by pairing with another product called Watson Discovery.

 Watson Assistant helps you to build chatbots and virtual assistants for a
variety of channels, including mobile devices, messaging platforms, and
even robots.
 Watson Assistant combines machine learning, natural language
understanding, and integrated dialog tools to create conversation flows
between your apps and your users.
 We offer package called Customer Care Voice Agent, which combines
Watson Assistant with speech to text, text to speech, and IBM voice agent.
These products can be easily integrated to enable Watson as an IVR.

Lesson Summary

In this lesson, you have learned:

  Watson Assistant helps you to build chatbots and virtual assistants for a
variety of channels, including mobile devices, messaging platforms, and
even robots.
 Watson Assistant combines machine learning, natural language
understanding, and integrated dialog tools to create conversation flows
between your apps and your users.
 https://www.ibm.com/cloud/watson-assistant/

Watson Discovery

Watson Discovery Service is a new suite of APIs that enables you to ingest,
normalize, enrich, index, search, and explore, your data, both structured and
unstructured.

Here's what you should know: Watson Discovery gives you the ability to, store,
analyze, and understand your data at scale.

First, Watson ingests and standardizes all your data, no matter the type. Just
load it and go.

Watson then enriches your data using natural language understanding

capabilities. Think sentiment and emotion analysis, named entity extraction,
and concept tagging.

But that's just the beginning, an easy-to-use query API gives you new ways to
segment and serves through your data.

For example, extracting sentiment from the past few months of your
customers' chat history.

You're not just getting Natural Language Processing with Discovery, you're
getting a host of out-of-the-box enrichment algorithms to find the signal and
your data. You can plot time series, perform aggregations, correlate variables.

With Discovery, developers can rapidly ingest and enrich data to build
cognitive cloud-based applications that connect proprietary structured and
unstructured data. Rest APIs make it straightforward to add cognitive apps
built with Discovery to any existing front-end solution. Watson Discovery
Service is available on the IBM Cloud, as well as other platforms. You can
leverage Watson SDKs, Web-based tooling, and Watson Discovery service
APIs to bring cognitive insights to your custom applications. Watson Discovery
allows you to use pre-built collections, such as Watson Discovery News and
your previously built custom collections. Discovery allows you to interrogate
your data in either a natural language format or a query language that will
search for both Boolean conditions and for phrases. Once the data is found,
Watson Discovery uses machine learning algorithms to order the answers so
that the most useful answers are listed first. Discovery gives you a head-start
on analytics by providing time sheets, graphs, and aggregations.

Watson Natural Language Understanding

The Watson Natural Language Understanding service allows users to

understand the topics mentioned in text by analyzing keywords, concepts and
categories. With entities, they can extract the people, places and
organizations mentioned in text. Users can also understand the sentiment
people express including sentiments expressed towards a specific keyword,
entity or phrase. Plus, they can take the sentiment analysis one level deeper
and to take distinct emotions behind what people write, when knowing if text
is positive or negative isn't enough. Finally, develop custom models using IBM
Watson Knowledge Studio and use them with Watson Natural Language
Understanding to tailor your solution by extracting domain specific entities
and relations. After building these custom components, you can publish your
custom model and use it through Watson natural language understanding on
the Watson Developer Cloud. Watson Natural Language Understanding
becomes even more powerful when paired with other services from the
Watson Developer Cloud. Use IBM Watson Developer Cloud Vision Services
to add images and videos to your analysis or use our speech services to pull
insights out of transcribed data from audio files.

Watson Natural Language Understanding is implemented as a sophisticated

suite of APIs, that support a variety of languages in varying capacities. They
are highly scalable and support large data-sets. Data can be sent as plain text
or HTML or is URLs and the results are returned for the features that are
specified. Machine learning or data science expertise is not required to use
the APIs. The data is always maintained and controlled by the owner.

Watson Knowledge Studio

IBM Watson Knowledge Studio, you can teach Watson about your specific
interest area faster through examples without writing any code. What used to
be measured in years, now is measured in weeks. Instead of requiring PhDs in
artificial intelligence, experts in any field can train Watson. To accelerate
training, you can pre-annotate documents using Watson Natural Language
Understanding, specify rules, or import dictionaries. Then you can fine-tune
results to meet your requirements.

Knowledge Studio creates a machine-learning model based on your training.

Watson can then learn on its own and identify entities and relationships that go
beyond your training. You use this model to build Content Analysis
applications using Watson Discovery, Watson Natural Language
Understanding, or Watson Explore. Watson Knowledge Studio helps you build
a machine learning annotator or to extract information specific to your
business requirements. Watson Knowledge Studio has built-in collaborative
workflow, which allows multiple subject matter experts to work together,
contributing to better, more accurate, and more quickly developed solutions.

Watson Knowledge Studio is a single interface that can be used for the end-to-
end development cycle of a machine-learning annotator. It does not require
data science expertise or any programming. Today, Knowledge Studio users
have trained Watson to identify critical content to stop cybersecurity attacks
before disaster strikes, and understand software support requests, and track
down information about rare diseases. Across financial, legal, technical, and
environmental industries, applications using Watson Knowledge Studio can
identify new opportunities, trends, and threats.

Lesson Summary
In this lesson, you have learned:

 Watson Discovery unlocks hidden value in data to find answers, monitor

trends and surface patterns with the world’s most advanced cloud-native
insight engine.
 Watson Natural Language Understanding offers a suite of features for
text analysis extracting entities, relationships, keywords, semantic roles in
up to 13 languages.
 Watson Knowledge Studio provides you with a way to teach Watson the
language of your domain with custom models that identify entities and
relationships unique to your industry.
 https://www.ibm.com/watson/services/discovery/
 https://www.ibm.com/watson/services/natural-language-understanding/
 https://www.ibm.com/watson/services/knowledge-studio/

Watson speech to Text

IBM Speech to Text uses advanced statistical modeling techniques,

developed over decades of IBM Research and refined by ideas from cognitive
computing to transcribe both high quality and lower quality audio from a wide
variety of source material. It uses the technology behind Watson to
automatically determine the most accurate result for words and phrases and
present them with confidence scores and other metadata.

Watson Speech to Text is an API based service that is specialized for

converting human voice into text featuring a special data format. The data
that is returned includes not only the translated text, but also alternative
translations along with a competent score for each one of those translations.
Out of the box, the service can translate general utterances and works with
many commonly used phrases. If you need Watson speech detects to
understand less commonly used words and phrases, perhaps those specific
to your industry, the service can be trained to recognize domain-specific
terms. IBM provides access to several software development kits available on
GitHub for working with the Speech to Text service. Since the service is
hosted on the IBM Cloud it is scalable, allowing you to have multiple services
working in concert to translate very large numbers of speech into text. The
service is highly customizable. It can be trained through the API to recognize
many words and phrases that may be specific to your use case. Watson
Speech to Text also supports several ways to connect to the service with
either live streams or pre-recorded audio. Most importantly, all the data that
passes through this Speech to Text service is your data.

Watson Text to Speech

Watson Text to Speech service is designed to convert written text into natural
sounding audio in a wide variety of languages, dialects, and voices.

The Text to Speech service includes a powerful real-time speech generation

system. It's highly customizable, and you can control very minutely, the
cadence, emotion, and tone of the speech produced. Out of the box, the
speech engine is extremely accurate. Built to support a wide variety of use
cases, the service can synthesize across many languages and numerous
voices, including different genders and ages.

The Text to Speech service supports the International Phonetic Alphabet and
the IBM symbolic phonetic representation notation, to enable you to construct
very specific pronunciations for terms, phrases, and words, that may appear
in your use case.

Watson Text to Speech gives you very detailed control over the pronunciation
of the different words and phrases that pass through the service via a
powerful customization interface. The Text to Speech service provides a
REST API, which allows you to submit text for speech processing on an as-
needed basis. It also has a WebSocket interface to enable you to stream
audio output from the Text to Speech service in real time. There's also Cross-
Origin Resource Sharing or CORS support built into the Text to Speech
service. This means the service is accessible directly by any configured
server or application without having to pass through specific servers or proxy
servers. So that you can work directly with the Text to Speech service,
subscriptions include access to software development kits and sample code
available on GitHub. All these tools can be put together using the Speech
Synthesis Markup link, which allows you to incorporate notation into your
customization of the Text to Speech service. Finally, when using Watson Text
to Speech service, you always maintain control and ownership of your data.

Lesson Summary

In this lesson, you have learned:

 Watson Speech to Text converts audio and voice into written text for
quick understanding of content.
 Watson Speech to Text automatically transcribes audio from 7
languages in real-time, in real time or from recorded sources.
 Watson Text to Speech converts written text into natural-sounding audio
in a variety of languages, dialects, and voices.
 Watson Text to Speech enables you to construct very specific
pronunciations for terms, phrases, and words that may appear in your use
case.
 https://www.ibm.com/watson/services/speech-to-text/
 https://www.ibm.com/watson/services/text-to-speech/

Watson Language Translator

The IBM Watson Language Translator service enables application developers

to directly process translations using a simple standard REST API. Language
Translator service offers multiple domain-specific translation models.
In addition, for texts with very specific terminology and language, the service
empowers users with three levels of customization. Forced glossary
customization allows you to upload pairs of matching terms in the source and
target language. These pairs completely overrule the original translation
model. Parallel phrase customization allows you to upload a series of
matching phrases in the source and target language. This differs from the
force glossary customization, because it does not completely overrule the
original translation model. Corpus level customization allows users to upload
a plain text file, that contains a body of text in the target language, that is
related to what you are translating. A monolingual corpus helps make your
translations more fluent and human.

Watson Language Translator service features Neural Machine Translation,

which supports customization in 14 language pairs. There was also a
Document Translator service to translate documents from one language to
another while preserving file formatting and type. IBM respects that you own
your data.

Watson Natural Language Classifier

Natural Language Classifier is text classification made easy. It applies deep

learning techniques to make predictions about the best predefined classes for
short sentences or phrases, such as descriptions, online reviews, job
descriptions, or key skills listed in resumes. Use machine learning to analyze
text, and label, and organize data into custom categories.

Organizations use Watson Natural Language Classifier to analyze product

descriptions and organize products into categories on online shopping
websites, to sort job vacancies into groupings, or to analyze the language
used in reviews and categorize those into types.

With Watson Natural Language Classifier, you can categorize social media or
analytics data, sort volumes of written content, evaluate the results, and
identify key skills.

The tooling features help you to classify and analyze your data. Use
classification for intent detection, creation of custom classifiers with
confidence scores, chain classifiers, and to classify multiple phrases in one
API call. Use data updates to update your training data based on recent
classification results. Multi-intent classification helps you to extract intense
from text, as well as separating and grouping text at various levels by
changing multiple classifiers.

Natural language classifier is beneficial because it allows you to scale your

classified data through the tooling and API features. You can leverage
Watson deep learning as a service to train a natural language classification
model on your data at faster speeds and in larger quantities.

Lesson Summary

In this lesson, you have learned:

 Watson Language Translator translates text from one language to

another, enabling you to communicate with your customers in their own
language.  
 Watson Language Translator supports customization which can be
especially helpful for translating technical terms.
 Watson Natural Language Classifier uses machine learning to analyze
text, and labels and organizes data into custom categories.
 https://www.ibm.com/watson/services/language-translator/
 https://www.ibm.com/watson/services/natural-language-classifier/

Week 3
Watson Studio

Watson Studio, you can simplify your data projects with a streamlined
process that allows you to extract value and insights from your data to help
your business get smarter faster. It delivers an easy to use collaborative data
science and machine learning environment for building and training models,
preparing and analyzing data, and sharing insights, all in one place.

Watson Studio's easy-to-create visualizations and drag and drop code put the
power of database decision making into the hands of any member of your
organization, with no need for IT assistance. And if you need access to open
source tools, the environment offers some of the most popular and powerful
ones available.
Ensures data scientists can share assets and work to solve problems within
the system rather than starting from scratch every time a new issue arises.
And developers can use that efficiency to quickly dive into building machine
learning and deep learning algorithms.

Watson Studio can help you get started with modeling templates for areas
such as visual recognition, language classification, and other tools from IBM
Watson services. Because Watson Studio is seamlessly integrated with the
IBM Watson Knowledge Catalog and intelligent asset discovery tool, you can
transform data and models into trusted enterprise resources, and collaborate
with confidence without compromising compliance, security, or access
control.

Watson Studio provides many benefits for organizations, helping to infuse AI

into the business and drive innovation. You can train Watson Studio with
embedded AI services, including Watson Visual Recognition. You can
customize your models and deploy them as APIs or Core ML by using open
source tools like Jupyter, Notebooks, Anaconda, and RStudio. Watson Studio
supports most popular code libraries as well as no code visual modeling with
Neural Network Modeler.

For designing neural architectures using the most popular deep learning
frameworks. In Watson Studio, you can interactively discover, cleanse, and
transform your data using Data Refinery. It helps you understand the quality
and distribution of your data with built-in charts and statistics and provides
visualized results through interactive dashboards. Watson Studio includes an
intuitive drag and drop interface that enables a non-programmer to speed up
the model building process by visually selecting, configuring, designing, and
auto-coding neural networks. From development and training to production
and evaluation, Watson Studio tracks your models over time to ensure you
have the best performance for any given task, using the best solutions across
the entire life cycle of your machine learning models.

Watson Machine Learning

Machine Learning is a subset of AI that uses computer algorithms to analyze

data and make intelligent decisions based on what it has learned.
With Watson Machine Learning, you can use your own data to create, train,
and deploy machine learning and deep learning models. Watson Machine
Learning gives computers the ability to learn without specific programming.

The Watson Machine Learning Service is a set of REST APIs that you can call
from any programming language to develop applications that make smarter
decisions, solve tough problems, and improve user outcomes.

Watson Machine Learning is integrated with Watson Studio, enabling you to

integrate AI into your applications and processes. You can create different
models, use dashboards to compare results, and run automated experiments
and self-learning models with the integrated Watson Machine Learning engine.

With Watson Machine Learning, you can create machine learning models
using visual modeling tools and quickly identify patterns, gain insights, and
make faster decisions. Use Jupyter Notebooks with Python, R, and Scala, or
work with R in RStudio to build sophisticated analytic models trained with your
own data.

Design complex neural networks, then experiment at scale to deploy optimized

deep learning models within Watson Studio. Watson Machine Learning brings
your organization the following benefits: It's easy and cost-effective to scale,
regardless of the environment you're working on, public, private, or hybrid
cloud.

It simplifies and automates much of the training and deployment process,

making your AI integrated products available more quickly.

And it comes with a wide variety of pre-trained models and open data sets.
Continuous feedback and performance monitoring help you to manage your AI
lifecycle.

Watson Knowledge Catalog

WKC provides a solution to the chaos of finding, understanding, trusting, and

sharing data, and the resulting loss of AI based business opportunities. IBM
Watson Knowledge Catalog is a collaborative enterprise asset catalog that's
designed to discover, and catalog enormous volumes of complex data and
analytic assets spread across multiple sources and provide advanced features
for collaborative work with data that helps to fuel new AI based business
growth. To help combat the issue of working with large amounts of data that's
spread out across the organization, Watson Knowledge Catalog can connect
to a diverse set of data sources, on-premises or in the cloud, including
RDBMS, NoSQL, file systems Hadoop, and more.

understanding and trusting your information is essential to delivering quality AI

and analytical results with confidence, Watson Knowledge Catalog uses the
power of Watson technology to enrich its metadata index of information with
classification, profiling, and quality assessment content, and provides
intelligent recommendations on the most relevant data assets that meets your
needs.

Watson Knowledge Catalogs active policy enforcement capabilities can control

access to information assets and support automatic and dynamic masking of
sensitive data elements that unlocks information, and safely powers
collaborative development of all data for business growth.

Watson Knowledge Catalog curates and catalogs all kinds of analytical and AI
assets including structured and unstructured data, notebooks, models,
dashboards, and more, and to support all forms of AI including machine
learning and deep learning while ensuring that data scientists can deliver
models efficiently. Watson Knowledge Catalog is tightly integrated with
Watson Studio, a single environment that manages the entirety of your data at
AI workflows.

Watson Open Scale

IBM Watson Open Scale, an open platform that gives businesses a clear
accurate view of their AI systems, helping them monitor and fine tune
performance across their lifecycle.

Traceability lets them retrace the process step-by-step right back to the
documents and data the AI drew from. If harmful bias has swayed an
outcome, a bias feature pinpoints how it occurred and automatically mitigates
it.

Watson Open Scale, a revolutionary technology called NeuNetS short for

Neural Network Synthesis, can quickly and automatically build models from
your latest data while recommending new models to be applied across your
business. Since Watson Open Scale is designed to be open, you can easily
connect new models.
IBM Watson Open Scale puts businesses in complete control of their AI's full
life-cycle, giving them the confidence to scale it and make fast accurate
decisions free from harmful bias, all while being open and easily plugged in to
existing processes.

Lesson Summary

In this lesson, you have learned:

 Watson Studio lets you build, train, deploy and manage AI models, and
prepare and analyze data, in a single, integrated environment.
 Watson Machine Learning enables you to use your own data to create,
train, and deploy machine learning and deep learning models.
 Watson Knowledge Catalog drives collaboration and transform data and
AI into a trusted enterprise asset through dynamic data policies and
enforcement.
 Watson Open Scale provides insights into AI health, recommends next
steps to improve outcomes, and orchestrates tasks to remediate issues
around performance, accuracy, and fairness.
 https://www.ibm.com/cloud/watson-studio
 https://www.ibm.com/cloud/machine-learning
 https://www.ibm.com/cloud/watson-knowledge-catalog
 https://www.ibm.com/cloud/watson-openscale/

Computer Vision

Computer Vision is one of the first machine learning fields all together to be
advanced.

The ImageNet challenge, about six or seven years ago when Geoffrey Hinton
from Toronto created the very first AlexNet neural network. And when he used
back propagation with a convolutional neural network for the very first time and
blew every other contestant in the ImageNet Challenge out of the water, that
was really one of the first real applications of machine learning. It really has its
roots in Computer Vision.

Computer Vision is another field that Watson excels in, both with your custom
models, your custom data, and with pre-trained models that Watson provides.
Applications for machine learning in Computer Vision, would be self-driving
cars.

Watson Visual Recognition

IBM Watson Visual Recognition. Visual Recognition understands visual data. It

can turn piles of images into organized information.

Watson Visual Recognition is an image recognition service that enables users

to quickly and accurately tag, classify, and train visual content using machine
learning. The service includes tooling that utilizes built-in capabilities, such as
the ability to quickly classify images, as well as the ability to build custom
classifiers using the IBM Watson Studio. Watson Visual Recognition models
can also be incorporated into Apple iOS mobile devices by using Apple's
CoreML framework. Flexible customization means that the service can be
trained with custom classifiers and connected to other services all while
maintaining privacy and security. The service is specialized for analyzing the
visual content of images or frames for videos. It is very fast and very scalable.
If you have a lot of images to analyze in a short period of time, the Visual
Recognition service can be scaled to handle heavy workloads. The Visual
Recognition service generates keywords describing what it sees in images
and includes number of features such as the ability to identify thousands of
different common objects that are encountered in video frames or still images.
The service also can detect human faces and identify a multitude of qualities
such as gender, and age or facial features. There are many other models
available for objects and categories such as food, text, and explicit images.

Custom models for Visual Recognition are created using the IBM Watson
Studio Custom Models and can be incorporated into Apple mobile devices
using Apple's CoreML framework.

Lesson Summary

In this lesson, you have learned:

 IBM developed some of the earliest computer vision systems.

 Some of the applications of computer vision include self-driving cars and
analyzing difficult to reach cabling for maintenance needs.
  Watson Visual Recognition allows you to quickly and accurately tag,
classify and train visual content using machine learning.
 https://www.ibm.com/watson/services/visual-recognition/

Watson Personality Insights

The Watson Personality Insights Service enables organizations to understand

personality characteristics, needs, and values, all from texts written by
individuals.

With this service, you can gain insights from an individual's social media,
enterprise data, or any digital data. You can tailor a customer's experience
based on how they write using that writing to tailor their experience to their
particular needs, wants, or personality. With the service, you can develop a
detailed personality portrait based on the Big Five Personality Model.

The Personality Insights Service allows social listening, where you can get
both the semantic meaning as well as a deeper analysis of your customer's
text.The service is based on the Big Five Personality Model, which is a
scientifically validated model of human personality.

Watson Tone Analyzer

IBM Watson Tone Analyzer enables organizations to understand emotions

that are being conveyed in text. Tone Analyzer uses linguistic analysis to
detect three types of tones from communications; emotion, social, and
language. That is emotions like anger, disgust, fear, joy, and sadness.

Social propensities such as openness, conscientiousness, extroversion,

agreeableness, and emotional range and language styles, for example,
analytical, confident, and tentative.

One of the features of Watson Tone Analyzer is a general purpose endpoint

where you can feed texts communication such as email, chat logs, text
messages, etc., into the service which will provide an analysis of the emotions
and tone being conveyed in those general purpose communications. It also
has a specialized customer engagement endpoint that can be used for
analyzing customer service interactions.

The Tone Analyzer can also be used to measure the performance of your
customer service agents. You can use it to enhance e customer service,
allowing you to monitor the reactions of your customers to provide them better
quality service.

The Tone Analyzer allows you to analyze sentence level text. In other words,
you don't have to have a whole lot of text to analyze and get meaningful
results.

The Tone Analyzer is accessible via a REST API, which allows a very easy
integration with many other services, including many of the services that
Watson

Lesson Summary

In this lesson, you have learned:

 Watson Personality Insights predicts personality characteristics, needs

and values through written text.
 Watson Personality Insights helps you understand a customer's
consumption preferences so you can tailor your social messaging for
each client based on their personality, needs and values.
 Watson Tone Analyzer understands emotions and tone in even small
amounts of text.
 Watson Tone Analyzer can be used to analyze the tone of online
reviews, enhance customer service interactions and improve customer
communications with chatbots offers, such as Watson Assistant, Speech
to Text, or Text to Speech.
 https://www.ibm.com/watson/services/personality-insights/
 https://www.ibm.com/watson/services/tone-analyzer/

Watson compare and comply

Watson Compare and Comply allows you to rapidly gather insights on the
documents in your company, particularly when it comes to contracts. Select a
contract to process and examine. Watson Compare and Comply associates
with each element, effectively a sentence within the document, with several
pieces of data and allows you to rapidly find elements of interest within the
document based on that data. First is the set of categories associated with the
element. The categories define the subject matter of the element.
Lesson Summary

In this lesson, you have learned:

 Watson Compare and Comply streamlines contract workflows to save

time and improve accuracy and simplify contract governance.
 Watson Compare and Comply identifies elements in each document,
and associate’s categories, and parties with those elements so that you
can analyze the relationships between them.
 https://www.ibm.com/cloud/compare-and-comply
 Congratulations! You passed!
 TO PASS 60% or higher
 Keep Learning
 Retake the assignment in 7h 56m
 GRADE

 90%

 Graded: More Watson AI Services

 LATEST SUBMISSION GRADE

 90%
 1.Question 1
 Which of the following built-in technologies helps you build machine learning models with a
simplified graphical interface?

 1 / 1 point

 Neural Network Modeler

 Data Refinery

 Watson Knowledge Catalog

 Intelligent Asset Discovery tool

 Correct

 Watson Studio’s built-in Neural Network Modeler helps you build models with a simplified
graphical interface.

 2.Question 2
 Which of the following statements about Watson Machine Learning is correct?

 1 / 1 point

 Watson Machine Learning is a set of REST APIs that you can call from any programming
language.

 Watson Machine Learning, does not allow you to use your own data to create, train, and deploy
machine learning and deep learning models.

 Watson Machine Learning connects only to IBM Cloud data stores, NOT to third party ones.

 Watson Machine Learning uses customizable proprietary tools to help you focus on your data.

 Correct

 Watson Machine Learning is a set of REST APIs that you can call from any programming
language to develop applications that make smarter decisions, solve tough problems, and
improve user outcomes.

 3.Question 3
 Which of the following statements about Watson Knowledge Catalog active policy enforcement
capabilities is correct?

 1 / 1 point

 Do not control access to information assets.

 Support automatic and dynamic masking of sensitive data elements.



 Lock down access to third-party development tools.

 Discover and catalog enormous volumes of complex data and analytic assets spread across
multiple sources.

 Correct

 Watson Knowledge Catalog active policy enforcement capabilities support automatic and
dynamic masking of sensitive data elements.  

 4.Question 4
 Which of the following statements about Watson Visual Recognition service is NOT correct?

 1 / 1 point

 Comes with several pre-trained models, like food recognition and object recognition.

 Provides custom training functionality so you can provide your own custom image data.


 Integrates with other tools in the Watson portfolio, such as Watson Studio.

 Comes trained to your custom specifications with the image data you supply to IBM.

 Correct

 Although you can train Watson Visual Recognition with your own custom image data, it only
comes with several pre-trained models.

 5.Question 5
 You are building a visual recognition system that will identify malformed widgets. Which
capability of Watson Visual Recognition will make this possible?

 1 / 1 point


 The ability to build custom models to recognize between good and defective widgets coming out
of the assembly line.

 The built-in models, such as facial recognition.

 The ability to create moving video from still images.

 The ability to incorporate Watson Visual Recognition models into Apple iOS mobile devices by
using Apple's Core ML framework.

 Correct

 Watson Visual Recognition can build custom models to recognize images specific to your
domain, which will allow you to train it to recognize malformed widgets.

 6.Question 6
 Which of the following statements about Watson Personality Insights is correct?

 0 / 1 point

 Watson Personality Insights supports only American English.

 Watson Personality Insights requires no training, it works right out of the box.

 Watson Personality Insights service tailors a customer's experience based on the tone of their
voice.

 Watson Personality Insights uses facial recognition to detect emotion of participants in video
calls.

 Incorrect
 Watson Personality Insights does NOT use facial recognition to detect emotion of participants in
video calls.

 7.Question 7
 You are building an AI system that can escalate a customer service call if the customer’s mood
becomes angry. Which Watson Service(s) can help you with this?

 1 / 1 point

 Watson Text to Speech

 Watson Language Translator

 Watson Tone Analyzer

 Watson Discovery

 Correct

 Watson Tone Analyzer allows you to monitor the reactions of your customers to provide them
better quality service. You can use it to escalate customer conversations if the customer gets
angry and the customer service agent cannot calm the customer.

 8.Question 8
 Which of the following statements about Watson Open Scale is correct?

 1 / 1 point

 Watson Open Scale is an open platform that operates and automates AI across its lifecycle.

 Watson Open Scale is used to manage large quantities of data.


 Watson Open Scale connects to open source translation services to translate large amounts of
text.

 Watson Open Scale does NOT provide traceability for the recommendations made by AI

systems.

 Correct

 Watson Open Scale is an open platform that gives businesses a clear accurate view of their AI
systems helping them monitor, automate, and fine-tune performance across their lifecycle.

 9.Question 9
 You are building an AI system that will help you to identify the payment terms specified in your
contracts with third-parties. Which of the following Watson Services could you use?

 1 / 1 point

 Watson Language Translator

 Watson Visual Recognition

 Watson Assistant

 Watson Compare and Comply

 Correct

 Watson Compare and Comply identifies elements within documents in your company and is
especially useful for analyzing the content of contracts. You can easily move between identified
elements, for example, to find all the obligations related to the category of payment terms.

 10.Question 10
 You are building an AI system that will help you to create personalized marketing content which
includes written and voice communication in an appropriate tone for each customer. Which
Watson Services can help you with this task?
  1 / 1 point

 Watson Tone Analyzer

 Watson Text to Speech

 Watson Visual Recognition

 Watson Knowledge Catalog

 Correct

 Watson Tone Analyzer allows you to personalize marketing communications with clients,
enabling you to most effectively communicate with clients based on their current emotional state.

Week 4

Lesson Summary

In this lesson, you have learned:

 The industries where Watson AI services are being used are many and
varied, including oil & gas, elevator maintenance, healthcare, automotive,
and many more.
 There are three broad categories of use case: Conversational
assistance, Compliance, and developing AI powered apps in-house.
 ADNOC are using AI to gather and retain knowledge from their
experienced workers, and to make it easily accessible to younger workers.
 Live Person are collaborating with IBM to create Live Engage, a single
platform that integrates AI, chatbots, and human agents.  
 Connie is a member of Hilton’s concierge team, providing natural
language assistance to hotel guests.
 Bradesco, one of Brazil’s largest banks, is using AI to improve response
time to customer queries.
 Coca-Cola Company are using Watson Services for Core ML to provide
their field engineers with a handheld repair assistant for beverage
dispensing machines.
 KONE are using IoT sensors in elevators to provide information for an AI
system that gives their maintenance team up to the minute information.
 Woodside, Australia’s largest oil and gas company is using AI to gather
information from their retiring workforce and make it easily available for
their workers.   

Note! You can upload documents in the form of PDF, HTML, JSON, WORD,
Excel, PowerPoint, PNG, TIFF, JPG. Any of the above filetypes for up to 50
megabytes.

After your document has finished processing, click on the "Schema" tag
highlighted by the red box, and click on document view. As you scroll down
you can see the sentiment level of each entity. In this case it is positive for
IBM Watson, IBM Services, IBM Cloud, and AI. So, this tool has improved my
workflow by providing me the sentiment for IBM Watson, IBM Services, and
IBM Cloud without someone reading through the whole document.

You can also extract important information from your document by using
keywords. Click on the Search tab and enter your keywords in the Search for
documents section. In this case we are searching for "Intellectual Property",
Watson Discovery has returned the result for "Intellectual Property" in our
document.