One of the main features of social robots is the ability to communicate and interact with people as partners in a natural way. However, achieving a good verbal interaction is a hard task due to the errors on speech recognition systems, and due to the understanting the natural language itself. This paper tries to overcome such kind of problems by presenting a system that enables social robots to get involved in conversation by recognizing its topic. Through the use of classical text mining approach, the presented system allows social robots to understand topics of conversation between human partners, enabling the customization of behaviours in their accordance. The system has been evaluated in different contexts, taking in account the quality and accuracy of the speech recognition syestem used by the social robot.
Anzalone, M.A., Yoshikawa, Y., Ishiguro, H., Menegatti, E., Pagello, E., Sorbello, R. (2013). A Topic Recognition System for Real Word Human-Robot Conversations. In Intelligent Autonomous Systems 12 (pp. 383-391). Springer Berlin Heidelberg [10.1007/978-3-642-33932-5_36].
A Topic Recognition System for Real Word Human-Robot Conversations
SORBELLO, Rosario
2013-01-01
Abstract
One of the main features of social robots is the ability to communicate and interact with people as partners in a natural way. However, achieving a good verbal interaction is a hard task due to the errors on speech recognition systems, and due to the understanting the natural language itself. This paper tries to overcome such kind of problems by presenting a system that enables social robots to get involved in conversation by recognizing its topic. Through the use of classical text mining approach, the presented system allows social robots to understand topics of conversation between human partners, enabling the customization of behaviours in their accordance. The system has been evaluated in different contexts, taking in account the quality and accuracy of the speech recognition syestem used by the social robot.File | Dimensione | Formato | |
---|---|---|---|
IAS_12_1_SORBELLO.pdf
Solo gestori archvio
Dimensione
3.04 MB
Formato
Adobe PDF
|
3.04 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.