Implementation of the inference engine based on molecular computing technique

Presents a novel application of molecular computing to problems of expert systems. A biochemical reaction on DNA strands is used to realise the backward chaining algorithm. The knowledge base for the expert system is based on the sticker model: memory strands represent rules while stickers represent facts and hypotheses. A series of experiments has been conducted to confirm the practical utility of this approach. In these experiments, the parameters of biochemical reactions were varied to determine true/false recognition accuracy. In addition, we discuss the fundamental issues of inference engines and try to enhance the physical insight into the dominating features of the approach proposed.