Artificial Conversational Entity with Regional Language

Abstract

During the lockdown period, people suffered a lot with much misinformation and conflicts regarding the hospitals and the dangerous COVID infection. A large number of people have common and simple doubts regarding the infection, quarantine, bed availability, oxygen cylinder availability. So, a patient with minor issues does not need to reach the hospital in person to consult with doctors for simple and similar queries in that lockdown situation. To neglect these difficulties, the chatbots were proposed with an AIML platform to provide details about infection rate, information regarding treatments, decision making regarding hospital admissions. Artificial Conversational Entity otherwise called chatbots acts as the conversational agent or talkbot or chatterbot which has the capability of performing an intellectual conversation with a human. There are many chatbots built using Watson, Google DialogFlow, Keras Seq2Seq, Gradient Descent algorithm, Beam search decoding methods, etc., Even though many chatbots respond in regional languages, not all the chatbots are appreciated for their efficacy and attractive customization. So, a chatbot with both the national language which is English, and the regional language which is the Tamil language, using the RASA framework has been developed. RASA NLU is well known for its efficacy and good customization methodologies. RASA NLU has a higher level of Application Programming Interface (API). So, the proposed method can be connected with a web page, and carousels are added to project the model very much attractive and deliver effective visualization. The carousels provide a wide range of information in regional language and it can be customized according to the user's requirements.