The Research on Identification Model and Related Regulation Strategies Selection for Forest Disease and Insect Pest

This paper deals with forest disease and pest identification expert system. From analyzing respective characteristics of forest diseases and forest pests, two corresponding identification modes were proposed, and several kinds of possibility in deducing regulation matching process are also analyzed. In the pest identification process, by combining the knowledge from taxonomy and visual characteristic together, a dynamic multiway tree for identification is built. Therefore, the user can choose the characteristics that familiar to them and decrease the retrieval times from knowledge base. In the disease identification process, a rewriting algorithm from text classification called optimal symptoms matching mode was presented with the related regulation strategy selections. Moreover, new thinking on the traditional regulation matching methods of expert system was summarized and the related building-up scheme and test results were also introduced in the present study. From the test results we found that optimal symptoms matching mode can simplify the diagnostic process and reduce the inference period.