Discoursive intention in machine interpreting systems: a bibliographic study

• Autor(a) principal: Esqueda, Marileide Dias
• Data de Publicação: 2019
• Outros Autores: Freitas, Flávio de Sousa
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Domínios de Lingu@gem
• Texto Completo: https://seer.ufu.br/index.php/dominiosdelinguagem/article/view/45057
• topic: Interpretação Automática; Sistemas de Interpretação Automática; Intenção discursiva; Machine Interpreting; Machine Interpreting Systems; Speaker’s Intention

Machine Interpreting or Speech-to-Speech Translation is a new technology that converts spoken utterances from one language into another through three functionalities grouped in only one software: Automatic Speech Recognition, Machine Translation and Speech Synthesis (or Text-To-Speech) (FREITAS; ESQUEDA, 2017). In order to develop systems of this type, several researchers first sought to understand human speech, its characteristics, how it is processed by the human brain and produced by the speech apparatus (LEE, 2015). It was only after such investigations that it was possible to develop techniques and approaches to process, represent and reproduce human speech through computational systems. Given the inconsistencies of the first results, which represented very literal and erroneous interpretations, scholars argue in favor of an approach that takes into account the speaker's intention, alluding to the context and the communicative purposes of the dialogues (JEKAT; KLEIN, 1996). Thus, this bibliographic study seeks to explore the general characteristics and evolution of the discursive intention capture approach in machine interpreting software.