2. Chabot
Conceptual
Introduction
Chatbot’s allow us to have a new way to automate our audience
engagement in a more efficient way that is always in line with your
strategy or pre defined goal.
What ever your objective, be it delivery of news, updates, events
information, information about your business, services, knowledge
base, FAQs, product showcase, campaigns, contest tools.
Chatbot’s can be described as small pieces of software follow a
dialogue script in order to grant a sort off interaction with brands.
The greatest challenges when building a Bot are:
• UX speech conception of the BOT key words and answers;
• Integrations with the business, a good WebAPI with the business is
a need
3. Chabot
Block Diagram
Architecture
Chatbot’s Arhitectures are typically built considering the 4 big blocks, is
not mandatory that you have all the blocks described:
1. Channels: the place where Chatbot interact (i.e Facebook
Messenger, Skype, Telegraph, Slack etc);
2. The Chat Bot algorithm ;
3. The business services that allows the chatbot to interact with the
Business logic (i.e a british airways Bot would need to had access
in real time to flights information to inform the user );
4. A IA Module that allows to create events associated with language
processing and associated integrations with the business;
1 2 3
4
Bot
Code BusinessWebAPI
Natural Language Processing
4. Chabot
Base comcepts
To the Channels we typically named as Bot Platform
To the base development environment for the Bot we give them the
name of Bot Framework
Bot Framework Bot Platform
Set the predefined functions and
some levels of abstractions related
to the base bot platform
Ecosystems where bots can
interact with humans
Define a programming language
for bot development
Allow users to find bots as a service
suppliers
Grants the build a Natural
Language Programming base
model
5. Must used
Frameworks
Microsoft Bot Framework: Offers a complete development
framework. All architecture blocks are exposed as
websirvices Rest.Allows to extend with .NET code. Comes
integrated with LUIS a natural language processing.
Demands an Microsoft Azure Deployment
Wit.AI: is the bot framework of Facebook. Allows easy
extraction off some user data.Very focosus on Facebook
platform
6. Chatbot
implementation
Chalanges
I. Well defined channel or bot platform target
II. Understant the possible customers journey when interacting with the
Bot
III. Choosing the language and culture for the Bot
IV. Train the base vocabulary considering the intents and choosen
journey
V. Define the needed information from the brand, and try to get that
information to the bot usingWeb Services integrations
VI. Study the target platform and understand what is needed to publish a
bot in that platform ( I.E Facebook demands an app creation and a
base page )