Xiong Fanlun

Study and implementation of the expert system for greenhouse tomato planting

In order to improve the decision-making capability of the classical expert system of agriculture, an integrated mechanism of the agricultural expert system and the virtual plant growth model simulated the plants figure is proposed by employing the idea of software integration. Making use of the virtual plant growth model, the decision-making capability of classical expert system can be improved and the decision result can be vividly expressed. The total structure, the modules, the construction of knowledge database, and the built process of the virtual tomato growth model of the greenhouse tomatos expert system are discussed in detail. Finally, an example of the systems application is presented.

Measures of fuzziness and entropy of fuzzy information

Vagueness of knowledge results from the imprecision and uncertainty of knowledge. In fuzzy theory, much attention has being paid to the measure of fuzziness of a fuzzy subset, while entropy, as a measure of uncertainty, plays a significant role in the field of information theory. The paper discusses, when there is or not a probability distribution on the non-empty definite universe U, the measure of fuzziness and entropy of a fuzzy subset and the conditional fuzzy entropy of a fuzzy subset, when there exists additional probability information and fuzzy information.

A Visualized Tool for Developing AIS

Abstract This paper presents a visualized tool : its design, implementation and application for developing agriculture intelligent system(AIS). Agricultural knowledge is represented and acquired in a visualized manner which not only makes the knowledge base easily to maintain but also enhances the transparency of the system and the convenience of operation. Besides an interface generator which is provided to do visual design of the lIser interface, visualization method is also used in complex task decomposition in order to generate heterogeneous problem solving agents for the final hybrid AIS which has synthesized problem solving strategies.

Two non-classic graphic knowledge representation methods

Abstract Knowledge representation is one of the main topics of the knowledge subject. This paper introduces two non-classic knowledge representation methods. These two methods are conceptual graph and CASNET model. The paper gives the formal definitions of these two non-classic methods. The implementations and the reasoning based on these two methods are discussed in detail. The knowledge representation ability of these two methods is well matched to classic methods. Finally, the advantages of these methods are also presented.

A research and application on relevancy of text vector model based on specific corpus

How to use the mass electronic information in Web pages, and how to process these information automatically and intelligently is a hotspot in current research. One key problem is how to effectively extract information hidden in Web pages. With the help of specific corpus, this paper studies the relevancy of Web data using the text vector model (TVM), we propose the calculation rules of text vectors, and retrieve data from the Web. We also introduce an application to retrieve agricultural information.

A Method for Deciding the Number of Hidden Neurons of the Feedforward Neural Networks

Abstract The number of hidden neurons of the feed-forward neural networks is generally decided on the basis of experience. The method usually results in the lack or redundancy of hidden neurons, and causes the shortage of capacity for storing infonnation or learning ovennuch. This research proposes a new method for deciding the number of hidden neurons based on decision-tree algorithm. Firstly, an initial neural network with enough hidden neurons should be trained by a set of training samples. Second, the activation values of hidden neurons should be calculated by inputting the training samples that can be identified correctly by the trained neural network. Third, all kinds of partitions should be tried and its infonnation gain should be calculated, and then a decision-tree for correctly dividing the whole sample space can be constructed. Finally, the important and related hidden neurons that are included in the tree can be found by searching the whole tree, and other redundant hidden neurons can be deleted. Thus, the number of hidden neurons can be decided. In the case of building a neural network with the best number of hidden units for tea quality evaluation, the proposed method is applied And the result shows that the method is effective.