Conversational Agent using Sequence to Sequence Model
Abstract
To create a digital conversational assistant that responds accurately and promptly to the users with the use of semantic and effective responses. The virtual assistant must give grammatically correct responses, which should have good proximity to the human language. In our scenario, there will be a conversational agent (abbreviated as CA) and the user who can interact with the CA directly in the form of textual queries. And the CA must infer the intent and generate an appropriate response based on the query’s intent. Though there are a lot of CAs that are already existing that serve the purpose, but the problem with the existing systems is that they sound more like machines than humans. They might solve the purpose of responding to the user, but that’s not the scope of our project. Our chatbot aims at generating responses to the user’s query which are semantic, grammatically correct and have high proximity to the way we converse with each other.