Introducing Refinement Operators into Inductive Metalogic Programming.
-
- YAMAMOTO Akihiro
- Department of Intelligence Science and Technology, Graduate School of Informatics, Kyoto Univeristy
Bibliographic Information
- Other Title
-
- 精密化演算子を用いた帰納メタ論理プログラミング
- セイミツカ エンザンシ オ モチイタ キノウ メタ ロンリ プログラミング
Search this article
Abstract
Inductive Metalogic Programming means Inductive Logic Programming with metalogic programming. In Inductive Logic Programming, all of training examples, data, background knowledge and hypotheses for inductive inference are represented in the form of logic programs, but inference mechanisms are not always in the form. Inductive Metalogic Programming is the case that inductive inference mechanisms are also represented in logic programs. Since such an inference mechanism takes logic programs as training examples or data and returns logic programs as hypotheses, the logic program representing the mechanism must be founded on metalogic programming. In this paper we point out two problems on the formalization of Inductive Metalogic Programming given in a previous work. We solve one of the problems by introducing refinement operators for hypothesis space, and the other by re-formalizing inductive inference in the new logic, proposed by Martin et al., for unification of induction and deduction.
Journal
-
- Computer Software
-
Computer Software 21 (5), 409-414, 2004
Japan Society for Software Science and Technology
- Tweet
Details 詳細情報について
-
- CRID
- 1390282679713520256
-
- NII Article ID
- 130004549030
- 110003743187
-
- NII Book ID
- AN10075819
-
- NDL BIB ID
- 7105887
-
- ISSN
- 02896540
-
- Text Lang
- ja
-
- Data Source
-
- JaLC
- NDL
- CiNii Articles
- KAKEN
-
- Abstract License Flag
- Disallowed
