Research Of The Knowledge Acquisition Methods In "Intelligent Potato Information Processing System"

Aiming at the present state of obscure characteristics of current agriculture knowledge and lack of agriculture experts etc., the developing of intelligent agriculture is becoming a inevitable trend. This paper takes the PAID4.0 Expert System as the platform to develop the " intelligent agricultural information process system " of National 863 project. The system is designed, which consists of two models i.e. the multimedia information and intelligent strategy, The adoption of the Browser/Web Server/Database structure makes the fulfillment of web resources share as well as the extension of application scope of potato information processing system.The method of knowledge representation based on rule extension is adopted in the knowledge representation mode. As a general rule, knowledge acquisition is the most difficult work of constructing Expert System, and it is the key and bottleneck of constructing Expert System. This paper explains the development of knowledge acquisition of National 863 project "Intelligent Agricultural Information Process System".The knowledge in knowledge base is acquired from two sources. One source is original knowledge. It is directly Stored from agriculture expert. Knowledge editor is the tool of this knowledge acquisition. This paper describes our design of knowledge editor of the system The knowledge editor is web application software using Asp and database technique. It has advantage of web application software. The client is multi flat roof, and knowledge engineer and agriculture expert can edit knowledge any time and anywhere. In this way, it can be more efficient.The other knowledge source is knowledge discovery and data mining. Considering there has been large amount of statistic-data about agriculture, using machine study to achieve knowledgeacquisition of automatic or Semi-automatic. This paper discussed the way to realize automatic acquisition using knowledge discovery in database and the data mining of association rule.The method of Subjective Bayes is adopted which made the consistence of the scope of reliability between user's answer and the method of rule based knowledge representation.In the design of system, the Artificial Intelligence, expert system, knowledge engineering and information technology are successfully connected. Meanwhile, the Comprehensive application of ASP, the information technologies, fuzzy reasoning.etc, ensures the organic unification of system's advance and practicability.