Chatbots

Chatbots is a rampant AI technology that are capable of interacting with humans in a more natural way. We could see multiple application of this technology across multiple sectors such as banking, healthcare, e-commerce and so on.

It is efficient in providing customer service to a certain extent, its service is available anytime anywhere, it saves money by employing an individual for providing services for the customers, hence by minimising the cost to serve.

Chatbots are capable of providing a conversational experience with real humans using natural language processing and AI. Chatbots are of four types namely – Scripted Chatbots, Intent Recognizers, Virtual Agents and Human-like Advisors.

Scripted Chatbots are chatbots that work on a pre-written script. When a user inputs a query the chatbot provides answer as based on the script. In such chatbots the query is written separately by using regular expressions and other strings. The chatbot looks for a certain keyword in the question and provide output. If the chatbot couldn’t find the keyword then the chatbot will output – “Sorry I don’t understand”

Intent Recognizers is a chatbot that accepts input in written or spoken form and classifies it based on the customer’s requirement. It is a type of Natural language processing (NLP) technique that is widely used in customer service providing or sales conversations.  

Virtual Agent is an AI software chatbot that works based on a set of scripted rules in order to provide automated guidance and services to humans. It is widely used in providing customer service and call centres.

Human-like Advisors are advanced AI chatbots that are capable of understanding and reason based on its environment as same as a human being. Compared to other chatbots it is quite advanced.

Human brain is capable of Learning, Reasoning, Perceiving, Understanding and interacting. Chatbots could mimic theses functions of human beings. For Learning a chatbot uses Reinforcement learning, reasoning it uses natural language processing, perceiving uses entity recognition, for interacting it uses Q&A scripts and dialog management and finally for understanding it uses Machine learning and Intent recognition.