A knowledge processing neural network based on automatic concept hierarchization

Masahiro Saito, Masafumi Hagiwara

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, we propose a knowledge processing neural network which is capable of inductive and deductive inference. The proposed network looks up relations between words in a concept dictionary and co-occurrence dictionary. First, the proposed network divides sentences into the subject words and the other words. Then these words are input into two-layer network. Second, hierarchical structure is composed using concept dictionary. Third, the network induces general knowledge from individual knowledge. We added a function to respond to questions in natural language with "Yes/No/I don't know" in order to confirm the validity of proposed network by evaluating the quantity of correct answers.

Original languageEnglish
Title of host publicationNeural Information Processing - 14th International Conference, ICONIP 2007, Revised Selected Papers
Pages539-548
Number of pages10
EditionPART 2
DOIs
Publication statusPublished - 2008 Oct 24
Event14th International Conference on Neural Information Processing, ICONIP 2007 - Kitakyushu, Japan
Duration: 2007 Nov 132007 Nov 16

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 2
Volume4985 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other14th International Conference on Neural Information Processing, ICONIP 2007
CountryJapan
CityKitakyushu
Period07/11/1307/11/16

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'A knowledge processing neural network based on automatic concept hierarchization'. Together they form a unique fingerprint.

  • Cite this

    Saito, M., & Hagiwara, M. (2008). A knowledge processing neural network based on automatic concept hierarchization. In Neural Information Processing - 14th International Conference, ICONIP 2007, Revised Selected Papers (PART 2 ed., pp. 539-548). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4985 LNCS, No. PART 2). https://doi.org/10.1007/978-3-540-69162-4_56