Problem-solving method with semantic net based on DNA computing in artificial intelligence

Abstract

Semantic Net is among the problem solving systems in artificial intelligence fields. In this paper, we demonstrate how to design DNA-typed Semantic Net in order to apply DNA computing to artificial intelligence. Moreover, we propose a problem-solving method with DNA-typed Semantic Net. In this method, it is possible to reason out a reference object by using DNA computing algorithm. Proposed DNA-typed Semantic Net is used as a molecular knowledge based system. Vertexes and edges of the DNA-typed Semantic Net are encoded to four kinds of nucleotide. Single-stranded DNAs are hybridized and ligated to let them the double-stranded DNAs with the complementary sequences of input molecules and knowledge based ones. For the molecular knowledge based system, we estimate the computational complexity by using a simulation. Proposed problem-solving method is performed by DNA-based computer for a future generation of artificial intelligence.