Natural intelligence - Commonsense question answering with conceptual graphs

Güler F. M., Birturk A.

18th International Conference on Conceptual Structures, ICCS 2010, Kuching, Malaysia, 26 - 30 July 2010, vol.6208 LNAI, pp.97-107, (Full Text) identifier

  • Publication Type: Conference Paper / Full Text
  • Volume: 6208 LNAI
  • Doi Number: 10.1007/978-3-642-14197-3_12
  • City: Kuching
  • Country: Malaysia
  • Page Numbers: pp.97-107
  • Keywords: Combinatory Categorial Grammar, Commonsense Reasoning, Conceptual Graphs, Open Cyc
  • TED University Affiliated: No

Abstract

Natural Intelligence (NI) is a question answering system based on Combinatory Categorial Grammar (CCG) as the theory of grammar and Conceptual Graphs (CG) for knowledge representation and reasoning. CCG is a lexicalized theory of grammar and very suitable for semantic analysis. Conceptual Graphs is a special kind of semantic network which can express full first-order logic. It aims to address the problem of commonsense reasoning in question answering, by using the state of the art tools such as C&C tools, Cogitant and Open Cyc. C&C tools are used for parsing natural language, Cogitant is used for Conceptual Graph operations, and Open Cyc is used for upper ontology and commonsense handling. © 2010 Springer-Verlag Berlin Heidelberg.