Transcrição de voz para texto e síntese de voz em chatbots

Abstract

The popularization of message exchange applications has changed the way people expect to interact with technology. Because of this, the use of chatbots has been gaining great prominence for presenting an interface similar to these applications. The Government of the State of Ceará seeks to make information about its services available to the population through the Service Charter, and the use of chatbots is a good interface to make this information more easily accessible. However, while the exchange of voice messages in message exchange applications is quite common and allows the inclusion of users with difficulties in communicating by text, most chatbots do not have this functionality, which ends up excluding a share of users. In view of this, this work seeks to create an interface capable of receiving and sending messages by voice, creating a data set and training a neural network model to produce a Ceará voice, to be integrated with the developed chatbots in the project of the Government of the State of Ceará: Digital Government and Big Data Platform to accelerate the Digital Transformation, which will be made available on different platforms, including message exchange applications.