Prototype for consultation of COVID information through a voice assistant

Abstract

With the advent of voice assistants, many of the activities that people do daily are more comfortable. Sending simple voice commands to attendees to perform predefined actions provides accessibility for people with disabilities who have difficulty manipulating technologies or applications. The data collection for this work focuses on the Covid-19 pandemic. The data sources are the National Service for Risk and Emergency Management and the Ministry of Public Health of Ecuador. A database is designed based on the collected data. A web application prototype uses the database and applies Speech to Text, Text to Speech, and Watson Assistant to provide a query service. The recognition of intentions and entities by Watson Assistant is used to respond simulating a conversation and with the recognized information structure the required query. This prototype aims to rapidly provide information on the Covid-19 pandemic through dynamically created queries using natural language input.