ChatBots: Understanding the What, the How and the Why?

Assisting a personal need or a business need has really become one of the biggest challenge in today’s fast and changing life. Few years before, we always had a person who would be an expert in some field and his/her advice would help us in some or the other way.

Nowadays, the experts are the same, but their knowledge has been transferred to machines and systems, which in turn provide us with the advice. It is really amazing and interesting to know, about how this works and how a machine can be an expert in order to give us an advice.

So let us see, how chatbots are trending in demand and how they are helping us in giving a fruitful output and some of the best advises.

THE WHAT: Definition and Example of ChatBots:

To start with the details on chatbots, let us first know what it is? So a chatbot is basically a computer program, which is designed in order to provide us with some output based on user’s input. The output generated may be in speech or textual manner. The chatbot when replies an answer to the user query, the user is unaware about the fact, that the computer or the system is answering his/her questions. Yes it is strange. Just think that you ask a question to your computer, and the computer gives you an answer based on the information it possess. That would be amazing. But systems do not have common sense.

The principle of chatbots is based on the Turing Test. The Turing Test was introduced by the scientist Alan Turing in the year 1950. Since that year till today research is still going on and things are improving each day in order to improve the user experience. The Turing Test states that: “It is a test for intelligence in a computer, requiring that a human being should be unable to distinguish the machine from another human being by using the replies to questions put to both. If the user fails to identify the Turing Test holds True and if not then it is False.”



This principle is applied to the concepts of chatbots, which are highly increasing for commercial usage in the industry. Consider an example of how chatbots work:

  • User: Hello.
  • Chatbot: How may I help you?
  • User: Can you tell me what is famous in India?
  • Chatbot: Why not? India is famous for its rich culture, its values and morals, secularism and democracy.
  • User: Oh! Great. Thank you for the information.
  • Chatbot: No problem. Need any further information?
  • User: Thank you.
  • …. (And so on the conversation continues between the user and chatbot).

THE HOW: Working of ChatBots:

So that was the definition and example about the chatbot and on what principle it worked. Let us now see, how the actual working takes place. As the computer system only understands 0s and 1s, so how can the system understand user’s speech and text information? This is important to know.

What chatbots do is, they extract some specific keywords from the user’s input query and then accordingly displays an output to the user. For example, you have asked “How is the weather today?” In this case, the chatbot will extract the keyword “weather” and based on this keyword, the expected responses of the bot can be as follows:

  1. Response1: “The climate is pleasant and good.”
  2. Response2: “Do you wish to plan a holiday in such weather?”
  3. Response3: “Do not travel. The weather is not suitable for travelling.”
  4. .. (Many other types of responses, based on the knowledge provided to the chatbot).

The very first chatbot developed was ELIZA. It was a weak artificial intelligent which processed user’s query based on keywords extraction and accordingly displayed the output to the user.

Chatbots can be of two types:

  1. Based on a set of predefined rules. Weak AI Chatbots.
  2. Artificial Intelligence Chatbots (Agents). Strong AI Chatbots.

The first one (weak AI chatbots) are only based on rules. You gather the expertise knowledge, define rules for the same and feed it into the system, and your system will assist you thereafter.

In case of the second one (strong AI chatbots) many things are included such as: computer vision (the bots which can perceive human behavior), machine learning (the bots learning from the changes in the external environment and improve their knowledge base in real time) and many more.

ChatBots: How they work
Figure1

 

THE WHY: Uses of ChatBots and its Applications:

We saw the initial two cases and understood the basics of a chatbot system. Why such system are useful and till what extent it can be useful, let us now see that. The advantages and disadvantages of chatbots. Lastly I’ll be concluding the article with the applications of chatbots.

So here are the advantages and disadvantages of a chatbot:

Advantages:

  1. Saves Manpower.
  2. Cuts down the number of resources utilization.
  3. Chatbots are much faster when compared with traditional assistance and messaging service.
  4. Chatbots can perform multitasking.
  5. It’s a computer, a system assisting you, so there is time boundedness.
  6. They are always straight to the point, crisp and clear and what they say and what they mean.
  7. Chatbots can be web or application based and can be more generic to reach a large number of audience.

Disadvantages:

  1. Chatbots systems find it difficult to understand normal human language. (This occurs mainly due to difference in languages and way of speaking).
  2. It cannot complete complex actions and queries. If it is programmed to figure out complex solution, the cost increases and moreover the complexity increases to a larger extent.
  3. There is a human interaction. Based on the knowledge it contains, it may or may not be reliable. So a good chatbot, needs to have enough expert knowledge in order to be reliable and efficient.

Considering these, we move ahead with its applications. Chatbots are used in each sector with one motive to interact with customers. This motive is, in order to generate more leads, to help user in order to find something or get something, to keep the user updated with information, to advice the user on his/her query and many more.



Applications:

  1. To assist users in providing useful information
  2. To interact with users and showcase new services and products of an organization.
  3. To help users in their shopping habits and more.
  4. To advice users on their expenses, their accounts and many more.

ChatBots have a lot of application. Recently in the Mumbai BMC Elections, Operation Black Dot, an organization which helped the election commission in order to generate more voter turnout made use of a chatbot. Their chatbot assisted users to find their nearest voting booth, in order to vote for the right candidate, in the right sector in Mumbai, India.

So this was all about chatbots. Stay tuned for more IT updates and trends.

Do like, share and comment your views on the article. Enjoy Reading!

 


Article written and submitted by: Akshay Rakesh Toshniwal

Leave a Reply

Your email address will not be published. Required fields are marked *