Automatic Induction of Domain-Related Information: Learning Descriptors Type Domains

Abstract. Learning in complex contexts often requires pure induction to be supported by various kinds of meta-information. Providing such information is a critical, difficult and error-prone activity. This paper proposes an algorithm to automatically identify types from observations, and studies its performance and robustness.

• Type: Conference
• Year: 2004
• Where: ECAI
• Authors: Stefano Ferilli, Floriana Esposito, Teresa Maria Altomare Basile, Nicola Di Mauro
• Keywords: Artificial Intelligence, Complex Contexts, ECAI 2004, Error-prone Activity, Pure Induction