This document is for a live course on O'Reilly's e-learning platform.
You can find details on the course at the following link:
Link to course registration
The courses are an extension of Artificial Intelligence by Example by Denis Rothman:
The programs for the course run both on Windows and Linux (see Ubuntu VM below). You can view and download them from the https://github.com/Denis2054/Hands-on-Chatbots repository. Some programs will be developed and run on cloud platforms.
The programs are under the MIT License. THe programs are for educational purposes only. Thus do not use them before checking Google, Facebook and Twitter security policies. Also be sure to follow web and General Data Protection Regulations (GDPR). Take your time and explore each policy rule. It is a major part of a project. The more you respect privacy and security policies, the more your users and customers will trust you!
When creating Twitter, Facebook, for example, you will need to provide your terms of service and privacy policy. If you do not have one, there are plenty of examples online. Start by googling "terms of service form and privacy policy examples" or similar search keywords.
A VM for Oracle VM VirtualBox version 5.2.8 is available for the course. It contains the pre-configured packages for node.js, ngrok and Python 3.x VM download link Password: neo Once the file is downloaded -first run an anti-virus on the file -then go to "Files" and select import VM.
This pratical chatbot course for beginners covers both types of conversational UI’s, i.e., Chatbot NLP text and speech user interfaces, to reach through multiple devices and platforms. We will take a project-based approach to understand how these UIs are built and the best use cases for deploying them.
We will start by building a basic NLP Chatbot with the newly-acquired-by-Google Dialogflow and implement it on your website. At this point, you will have a customized text and voice Chatbot you can use for professional or personal use, ready on your website.
Moving on to an in-depth understanding of the Chatbot, we will explore leak management addressing the questions that were entered by a user and did not find an answer. We will learn how to bring the quality of the chatbot to a higher level.
Moving on to Facebook Messenger we will understand the basics of bot integration & see how to integrate Chatbot to social networks. We will explore text and speech UIs capable of interacting with digital personal assistants such as Google Assistant and Google Home. We will then run the applications on PCs and smartphones.
Finally, we move on to creating a Task model that can perform complex tasks such as ordering and planning events. We will use the Fulfilment functions of Google Dialog Flow to implement a web service, sending requests and waiting for responses. We will implement ready-to-use machine learning algorithms to enhance the services.
By the end of the course, you will be confident enough to create your line of Chatbots and speech UIs on the platforms we have studied. You will also be confident enough to continue your seamless integrations through Dialogflow’s cross-platform functionality on leading platforms such as Amazon Alexa, Microsoft Azure’s framework for Crotona or Skype, and more.
You will also have learned how to build a solid text/speech Chatbot standalone website version that you can deploy, update and maintain on several platforms seamlessly.
This training is aimed at teaching you to improve your processes by creating a fully operational Chatbot through Google Dialogflow that can be deployed on many platforms. You will be able to explore the ground-breaking combination of Chatbot Text/Voice UIs, event scheduling, and Machine Learning. You will be able to use this course to promote yourself through bots on your favorite social networks, enhance your company’s web services and improve the establishing of your personal brand.
● Create a Chatbot with a text and voice UI. Deploy it on your website during this course. ● Deploy cross-platform Chatbots along with services including machine learning functions on several social networks and home devices. ● Understand the principles and technical aspects of a successful, professional and personal Chatbot ● Create a complete NLP Chatbot that can call standard or machine learning services ● Explore personal assistant customized dialogs on Google Homes ● Publish your NLP Chatbot on your Facebook page, Twitter account, Skype and more
The course is aimed at developers, consultants, and IT Operations who want to enhance their Chatbot culture through Chatbot technology. If you have been struggling to find the time to gain proficiency and confidence with Chatbot, customized dialogs, and everyday automatized tasks, here is your one-stop solution!
Artificial Intelligence by Example
Note: For developers familiar with AI, go through chapters 8, 9, 11-17. For beginners in AI go through the book and programs to get the feel of AI. Focus on chapters 15, 16 and 17. However, it is important to understand the underlying concepts of AI to design an efficient text/speech Chatbot.
The main preparation is for you to think of how you would like to use a text/speech Chatbot on your website. What theme would you like the bot to learn to help sell or promote your products, services or personal brand?
Write down a few basic questions and answers you would like to start building your bot with. You will be implementing this during the live training on your website (you can turn it on and off during testing). Then you will be learning how to deploy it on several key platforms. With this in mind, the course will take you where you need to be to finalize your goals right after the course.
###Materials, downloads, or Supplemental Content needed in advance
· Github programs
· The material will be downloadable in real-time during the course. The size of the packages will remain small enough to that effect.
· Google Account
· Dialogflow account
· Chrome (just for the course to be sure; after the course, any browser that works well with JavaScript) with Chrome V8
Note: There is no need to install Node.js. We will use an online version installed in our Google Dialogflow environment.
The timeframes are only estimates and may vary according to how the class ises progressingse
The timeframes are only estimates and may vary according to how the class is progressing.
Section 1: Building and getting your Chatbot online on your website (Instructor Lecture,se lab, Q&A), duration: 1 hour
Lecture: How to create a fully operational basic Chatbot with Google Dialogflow. Run the Chatbot on as an operational service on your website from any machine or Smartphone. Chrome is recommended for this exploration. Any recent browser can be used beyond this first implementationse.
Explore the basic concepts of a chatbot and implement them on your website during the live session. Your newly created chatbot will also be able to run with an url, you can add to your social network page(s). Break: 10 misens
How to make your chatbot more human with many possible variations of the same dialog, small talk and leak detection. Leak detection functions will be explored through the chatbot's logs to find unanswered questions, mismatches, and errors.
Enabling Small Talk functions, learning how to detect leaks in the dialog through the log files and the query interface. Modify the NLP Chatbot to adapt to the user’s needs. Improve the dialog. Break: 10 mins
Section 3: The integration functions for Facebook, Twitter, Alexa and other platforms (Instructor Lecture, lab, Q&A), duration: 1 hour
Building a solid practical chatbot takes a lot of energy and time. We will explore how to use your chabot on multiple platforms through powerful integration interfaces that do not require writing code.
Lab: Explore an existing chatbot built on one platform , Google Dialogflow, on another platform, Facebook.
Modify the NLP Chatbot to adapt to the user’s needs. Improve the dialog. Break: 10 mins
Section 4: Customizing Google Assistant with a Chatbot (Instructor Lecture, lab, Q&A), duration: 1 hour
The different ways to run a customized NLP Chatbot with Google Assistant. Add exciting media to your dialogs. Learning how to run Google Assistant on all platforms and use it with smartphones (“Go Google”). Learning how to improve the NLP Chatbot through conversational traps.
Beginning to deploying a chatbot on Google Assistant and preparing it to apply for Google validation to be used on Android,se Google Home, and other platforms.
Section 5: Transform Google Home into a powerful personal assistant with a customized Chatbot (Instructor Lecture, lab, Q&A), duration: 1 hour
How to customize an NLP voice Google Assistant Chatbot for Google Home. You will see how the chatbot can then be exported for Amazon Alexa and other platforms if you wish once the course is completed.
Advanced examples will be provided for Google Home with real-life applications and explore the use of Google Assistant on Android Smartphones for both personal use and corporate use.
Lab: The chatbot will be explored through the Google Assistant simulator to understand how the actions work.
Break: 10 mins
Section 6: Let’s add services to a Chatbot through Dialogflow Fulfilment (Instructor Lecture, lab, Q&A), duration: 1 hour
We will explore how to use some of the advanced Dialogflow node.js JavaScript functions to customize the chatbot and link it to other websites and services. Note: Google Fulfilment uses an integrated version of node.js (nothing to install locally).
We will then using some of the advanced Dialogflow node.js JavaScript functions to customize the chatbot and link it to other websites and services. Break: 10 mins
Section 7: Let’s make the Chatbot smarter with machine learning (Instructor Lecture, lab, Q&A), duration: 1 hour
Lecture: Explore ways to include machine learning modules into an NLP Chatbot to transform it into a cognitive Chatbot. At this point, the Chatbot, with the services described in Section 6 and cognitive machine learning functions will begin to understand concepts better than humans in many areas.
Implementing an artificial intelligence in your chatbot through ready-to-use machine learning functions. Break: 10 mins
Section 8: Fully Autonomous Services that will help a Chatbot make decisions faster than a human in specific fields (1 Hour - seInstructor Lecture + Q&A)
Explore how to transform an NLP Cognitive Chatbot into a real-time decision-making tool for professional through a case study in the trend used by corporations such as Amazon’s real-time apparel manufacturing process. Discover how cognitive chatbots will totally## surpass humans in both services and manufacturing in the near future. The chatbot users will begin to view the chatbot as a vital personal asset. Explore how chatbots will become independent super-intelligent entities in the near future. Understand how AI connected humans will surpass unconnected humans.
Implementing a basic prototype for a decision-making chatbot manager that can replace an unavailable manager or an expert. The teamse can ask questions and obtain critical answers.
Interactive Discussion on the basic and ground-breaking aspects of the course
Denis Rothman graduated from Sorbonne University and l'Université Paris-Diderot, writing one of the very first word2matrix embedding solutions.
He began his career authoring one of the first AI cognitive Chatbot 30+ years ago applied to a cognitive & digitized language teaching Cse
##hatbot. He customized it for Moët et Chandon (LVMH) and scores of companies in various forms.[Article]
(https://www.linkedin.com/pulse/did-you-miss-ai-parsing-train-denis-rothman)
He has authored a profit orientated AI resource optimizing system written in Horn Clauses in Prolong for IBM and implemented in corporate environments. He also transposed it in C++, Java and presently in Python/Tensorflow.
In the years after, he authored an AI APS (Advanced Planning and Scheduling) solution based on cognitive patterns. This #AI software is used worldwide to this day in the aerospace, train, energy, apparel and many other corporate fields.
Name: Denis Rothman
Personal Websites and pages:
bio on Amazon
web site
LinkedIn
Blog