Create your first artificial intelligence chatbot from scratch

In this article, I will show you how to build a simple chatbot using python programming language. We will not use any external chatbot packages. The whole project will be written with plain Python. This is a great way to understand how chatbots actually work. Learning the behind scenes will also give us an insight of the chatbot packages. I will share some Chatbot building platforms in the introduction section of this article. Feel free to check them later if you are interested to learn more about Chatbots. I can’t wait to walk you through in this project, let’s get started!

Table of Contents:

• Introduction
• Step 1 – User Templates
• Step 2 – ChatBot Responses
• Step 3 – Respond Function
• Step 4 – Relation Function
• Step 5 – Send Message Function
• Final Step – Testing the ChatBot

Introduction

In 1994, when Michael Mauldin produced his first a chatbot called “Julia,” and that’s the time when the word “chatterbot” appeared in our dictionary. A chatbot is described as a computer program designed to simulate conversation with human users, particularly over the internet. It is software designed to mimic how people interact with each other. It can be seen as a virtual assistant that interacts with users through text messages or voice messages and this allows companies to get more close to their customers.

There are many ChatBot platforms that help companies to create personalized chatbots. The database and APIs play big role in these chatbot platforms. Some of those platforms are: Amazon Lex, Microsoft Azure Bot, ChatterBot. You can check their websites to learn more.

Hopefully in a future article, I would like to show how to create an advanced level chatbot using one of those platforms, but for today we will keep things simple and basic. If you are ready, let’s start programming!

Step 1 – User Templates

First thing first, we will tell the chatbot to ask the user’s name. This will make the our chatbot more realistic. Let’s run this line in our Jupyter Notebook:

print("BOT: What do you want me to call you?")
user_name = input()

Now, let’s create the templates. In this chatbot project, we will have only two people: the user, and the chatbot. When printing out the conversation timeline, these templates will help us to know who is saying what. The values that will go between the curly brackets will be responsive, you will understand what I mean in the next step.

bot_template = "BOT : {0}"
user_template = user_name + " : {0}"

Step 2 – ChatBot Responses

This step is a little long because it handles most of the conversation. We are defining the chatbot in this step. We have to put in the questions that our chat is expected to hear and the answers that the chatbot has. To make things more interesting, I’ve created three responses for each question. This way, the bot will have a different answer to the same question. More details will be under the code. Let me show the code first:

name = "Funny Bot 101"
weather = "rainy"
mood = "Happy"


responses = {

"what's your name?": [
   "They call me {0}".format(name),
   "I usually go by {0}".format(name),
   "My name is the {0}".format(name)
],

"what's today's weather?": [
    "The weather is {0}".format(weather),
    "It's {0} today".format(weather),
    "Let me check, it looks {0} today".format(weather)
],

"Are you a robot?": [
    "What do you think?",
    "Maybe yes, maybe no!",
    "Yes, I am a robot with human feelings.",
],

"how are you?": [
    "I am feeling {0}".format(mood),
    "{0}! How about you?".format(mood),
    "I am {0}! How about yourself?".format(mood),
],

"": [
    "Hey! Are you there?",
    "What do you mean by saying nothing?",
    "Sometimes saying nothing tells a lot :)",
],

"default": ["this is a default message"]

}

• In the beginning of this code, we are defining some variables that we want to use in the conversation. The great thing about this is that it can be reusable in the code. And when we want to change the value, we don’t have to go through all the line, changing the variable’s value will update the whole lines.
• As you can see, we are using “format()” in the responses. This function helps us to pass the values inside the strings very easily. The format is much clear this way. Also, you can pass one more than one variables using the format function. So there is no limit if you want to add more layered conversations.
• The responses are created as dictionaries. So the answers are found after knowing the question. To make things cool, I’ve even added some responses even when the user replies without writing anything.

Step 3 – Respond Function

Nothing complicated here. We are defining the function that will pick a response by passing in the user’s message. For this function, we will need to import a library called random. Since we don’t our bot to repeat the same response each time, we will pick random response each time the user asks the same question.

def respond(message):
   if message in responses:
      bot_message = random.choice(responses[message])
   else:
      bot_message = random.choice(responses["default"])
   return bot_message

Step 4 – Relation Function

This is an extra function that I’ve added after testing the chatbot with my crazy questions. So, if you want to understand the difference, try the chatbot with and without this function. You will see why I decided to write this function. And one good part about writing the whole chatbot from scratch is that we can add our personal touches to it.

In the Chatbot responses step, we saw that the chatbot has answers to specific questions. And since we are using dictionaries, if the question is not exactly the same, the chatbot will not return the response for the question we tried to ask. Sometimes, we might forget the question mark, or a letter in the sentence and the list can go on. In this relation function, we are checking the question and trying to find the key terms that might help us to understand the question. You will understand what I mean when look at the function.

def related(x_text):
   if "name" in x_text:
      y_text = "what's your name?"
   elif "weather" in x_text:
      y_text = "what's today's weather?"
   elif "robot" in x_text:
      y_text = "are you a robot?"
   elif "how are" in x_text:
      y_text = "how are you?"
   else:
      y_text = ""
   return y_text

I used if and else but it could be done using switch cases too. Feel free to convert it to a switch case.

Step 5 – Send Message Function

Almost done, this is the last step of us writing a function. In this function, we are writing the conversation in timeline order and also calling the response function that was created earlier. After getting the answer from response function, we are passing it into the conversation print outs. This is a short function and here it is:

def send_message(message):
   print(user_template.format(message))
   response = respond(message)
   print(bot_template.format(response))

Final Step – Testing the ChatBot

Great, you made it to the final step! Everything is ready. All we need is to get the program started. We will write a while loop and pass the functions inside it. I preferred using infinite while loop so that it repeats asking the user for an input.

while 1:
   my_input = input()
   my_input = my_input.lower()
   related_text = related(my_input)
   send_message(related_text)
   if my_input == "exit" or my_input == "stop":
      break

• Firstly, we are asking for an input from the user. We can think of it as our bot is listening to the user here.
• Secondly, we are lowercasing all the letters in the input. This way we don’t have to worry if the user used capital or lower case letters. Makes things better.
• Thirdly, we are running the related function. This way, we predict the question asked by checking the key words.
• Fourthly, we are sending the returned question to the bot using the send_message function that we created in the fifth step. The response function is also called inside send_message function.
• Lastly, I’ve added an if statement to check if the user wants to quit the conversation. Entering “exit” or “stop” into the input box will trigger the program to stop. Since we are in while loop, this also helps to break the loop.

Example conversation I had with my Funny Bot 101:

That’s all for this article. Hope to see you in the next one.

Check my Medium Blog and YouTube Channel to stay inspired. Thank you,

More articles by author:

Building a Face Recognizer in Python

Blending Two Images using OpenCV