Home // COGNITIVE 2014, The Sixth International Conference on Advanced Cognitive Technologies and Applications // View article

Training a Cognitive Agent to Acquire and Represent Knowledge from RSS feeds onto Conceptual Graphs

Authors:
Alexandros Gkiokas
Alexandra I. Cristea

Keywords: Cognitive Agent; Reinforcement Learning; Conceptual Graphs; Expert Systems; Imitation Learning.

Abstract:
Imitative processes, such as knowledge transference, have been long pursued goals of Artificial Intelligence (AI). The significance of Knowledge Acquisition (KA) in animals and humans has been studied by scientists from the beginning of the 20th century. Our research focuses on observational imitation through agent-user interaction, for acquisition of symbolic knowledge. The cognitive agent (CA) emulates an imitative learning system, trained for the purpose of learning to represent knowledge, acquired from Rich Site Summary (RSS) feeds. It learns to autonomously represent that knowledge in a manner that is both logically sound, and computationally tractable, through the fusion of Conceptual Graphs (CG) and Reinforcement Learning (RL). The novel algorithm enabling this agent extends Reinforcement Learning, by approximating de cisions via exploitation of distributional and relational semantics governing the knowledge domain.

Pages: 189 to 194

Copyright: Copyright (c) IARIA, 2014

Publication date: May 25, 2014

Published in: conference

ISSN: 2308-4197

ISBN: 978-1-61208-340-7

Location: Venice, Italy

Dates: from May 25, 2014 to May 29, 2014