Page 137 - Artificial Intellegence_v2.0_Class_11
P. 137
They do this so that they know how the chatbot should respond. These chatbots use very simple or complicated
rules. However, they cannot answer questions outside of the established rules. Also, these chatbots cannot
learn through interactions. They can only handle scenarios you are training them for. For example, a basic text
messaging application, such as the ones used by some retail organisations to provide coupon codes or offers, is an
example of a rule-based chatbot. At a local coffee shop, you could see a sign that says, "Text COFFEE to 12345 for a
10% off coupon."
When your phone's text messaging system transmits the word "COFFEE," the program examines it and recognises it
as a command. It then follows a rule created by the developer and returns the coupon code you requested almost
instantly.
However, if you send, to that same chatbot, "I want one of those coupons," it would have no idea what you are talking
about. It's just programmed to act in response to the word "COFFEE." Also, it will not understand if you misspelled the
term and texted “COFEE.”
Some of the advantages of rule-based chatbots are:
• Easy to train (less expensive)
• Very responsible and safe
• May contain some interactive elements and media
• Not limited to text interactions
AI-based Chatbot
By comparison, AI chatbots use machine learning to understand the context and intent of a question before
formulating an answer. Hence, these chatbots are also known as machine learning based chatbots. These chatbots
generate their own answers to more complicated questions using natural language responses. The more you use
and train these bots, the more they will learn and the better they will work with the user. For example, Kuki is an
AI-based chatbot.
Many people consider AI bots to be more sophisticated cousins of chatbots. They work well for companies that have a
lot of data. Although training takes longer in the beginning, AI chatbots save a lot of time in the long run.
AI chatbots:
• learn from the collected information.
• continually improve as more data enters the system.
• understand behaviour patterns.
• have a broader range of decision-making skills.
• can understand many languages.
GENERAL AI
IMPOSSIBLE
Conversation capability Open Domain Closed Domain EASIEST TO ACHIEVE HARDEST TO ACHIEVE
INTELLIGENT
RULE BASED
MACHINE
HARD TO ACHIEVE
Retrieving responses Generating responses
Response mechanism
Source https://dzone.com/articles/how-to-make-a-chatbot-with-artificial-intelligence
AI Applications and Methodologies 135

