A review of the current systems that aid in carrying out tasks traditionally performed by human assistants, their advantages, and drawbacks are presented.
In the current day and age, systems that aid in carrying out tasks traditionally performed by human assistants are becoming more popular as human operators need more time for executing queries such as booking a ticket, purchasing items, or obtaining services. A single request may comprise many inquiries for information available on the Internet. Since business performance places a premium on time efficiency, an alternate method of accepting requests must be considered. Chatbot provides support 7 days a week, round-the-clock, and is not constrained by limited working hours. We studied various papers which approached this topic and presented their approach on creating a chatbot. Majority of the studies indicated that a retrieval-based chatbot is the best way to classify the user's context in a discussion using Utilizing Natural Language Processing (NLP) to assess the query and get some relevant keywords. Intent recognition, name entity recognition (NER) tagging, and information retrieval are essential processes in NLP. part of speech (POS) tagging helps to parse the interpretation of chat text. The rule base is intended to gather all keywords based on conversation content and is language-specific. The terms used in the booking dialogue include the date, time, city, scheduling, rescheduling, or cancellation of an appointment. In this paper, we have presented a review of the current systems, their advantages, and drawbacks.