Natural Language Generation in the context of Multimodal Interaction in Portugu...

• Author(s): José Casimiro Pereira ; António J. S. Teixeira ; Joaquim Sousa Pinto
• Date: 2013
• Origin: Electrónica e Telecomunicações
• Subject(s): natural language generation; NLG; multimodal interfaces; fission; adaptability to users; data to text; portuguese; natural language generation; NLG; multimodal interfaces; fission; adaptability to users; data to text; portuguese
• Document Type: Article
• Language: English

Description

In many situations, systems and people have to interact and share information. With demographic ageing a group of users that deserves special attention are the olderadults. Multimodal interaction aims to improve the interaction between human users and electronic devices. A key technology for the output part of a Multimodal system isconversion of the concepts handled by systems into readable text or audible speech output, area with almost no work for Portuguese. As a starting point for the development of output generation modules for use in multimodal interaction contexts, we developed, as proof of concept, a prototype of a NLG module to support transmission of information ondrugs to older adults based on a data-driven translation approach. First sample results of the generation are presented. In many situations, systems and people have to interact and share information. With demographic ageing a group of users that deserves special attention are the older adults. Multimodal interaction aims to improve the interaction between human users and electronic devices. A key technology for the output part of a Multimodal system isconversion of the concepts handled by systems into readable text or audible speech output, area with almost no work for Portuguese. As a starting point for the development of output generation modules for use in multimodal interaction contexts, we developed, as proof of concept, a prototype of a NLG module to support transmission of information ondrugs to older adults based on a data-driven translation approach. First sample results of the generation are presented.