A REVIEW ON ARTIFICIAL CONVERSATIONS FOR CHATBOTS: ARCHITECTURE AND ALGORITHMS

Abstract

Chat bots, or conversational interfaces as they are also known, are a new way of interacting with computer systems for individuals. Traditionally, a question can be answered by a software program using a search engine or by completing a form. The existing approaches to the generation of artificial conversations enhance linguistic models. The existing approaches are carried out in situations that require contextual continuity over a number of statements. Among the algorithms like natural language understanding, ASR process model, restricted Boltzmann machine we have chosen natural language understanding, because it comes under natural languageprocessing which involves speech recognition, parsing, stemming, lemmatization and optical character recognition and thus increases accuracy where ASR process model focuses more only on speech recognition. Here, we provide a set of inputs and responses which is processed by the bot to give the customer an accurate response.