The Restriction of Qualitative Reasoning Result Based on Rough Set Theory

Abstract

From the analysis existing problems of Qualitative reasoning results, the necessity of qualitative constraints was put forward. Several existing Qualitative constraints were summed and a new qualitative constraints rough qualitative constraints was advanced, which uses rough set attribute reduction algorithm, reducing the results of qualitative reasoning. Rough Qualitative reasoning software of the overall design of the process and function modules were introduced. Taking Material Supply Support process to the results of qualitative reasoning as an example, results from the rough qualitative constraint reasoning, reasoning by important results for the material supply security decision-making reference. Keywords-Qualitative Rough set, Reduction, Reasoning results I. . THE PROBLEMS IN QUALITATIVE REASONING RESULT The qualitative simulation model Based on signed directed graph (Signal Directed Graph, SDG) can express complex causal relationship, with the ability to accommodate large-scale potential of information, particularly for the system dangerous identification and fault diagnosis, has good completeness, but its lower resolution, although the overall result cannot focus. Therefore, the results of reasoning need to go through filtrating, to get useful information. Studying material supply, through analysis of qualitative simulation model of SDG qualitative reasoning results to find the defects aspects in material supply support process, but the various elements of the different results produced by the impact of different factors, the simulation results produce a lot of redundant data[1]. Through filtering it found the important factors, which can provide the basis for decision-making material supply support. II. TYPES OF QUALITATIVE CONSTRAINTS In early the qualitative constraints including qualitative algebraic constraint and qualitative monotonic relationship between constraints. Qualitative algebraic constraint some basic algebraic and differential relations, of which +, * derivative is a symbol of four basic algebraic constraint, that the qualitative constraints, add, subtract, multiply, except, qualitative monotonic relationship that the incremental impact and the reduction between the two constraints[2]. In the process of Qualitative modeling, causality often defined as the property value of model boundary, some heuristic rules are often used as the conditions of a model deflection (deflection defined as the static model of SDG qualitative model of a qualitative reasoning leaving node by the state to increase or decrease the normal state). Qualitative constraints mathematical description is: probability qualitative qualitative constraints facing the Piecewise function, fuzzy qualitative rough qualitative constraints[2]. Probability qualitative constraint the use of a variety of qualitative information, and to introduce random elements, change the uncertain relationship to the conditions probability, so the formation of a Bayesian network. Piecewise function qualitative constraint based on QSIM algorithm, made two major improvements: an increase of constraints to determine TF, access will be bound to a tree structure, scope limitations as to increase the assignment constraint EQ, does not solve the function discontinuities. Fuzzy qualitative constraint a semi-quantitative qualitative constraint. In the fuzzy simulation algorithm, due to the introduction of fuzzy metric spaces, allows the use of fuzzy relations to represent the qualitative function the use of partial order information to establish quantitative functional dependencies between variables, Fuzzy qualitative simulation mentioned in the constraint not used to solve the unknown variable values, but to use the fuzzy relationship between the variables, testing the consistency of its value, a form of qualitative constraints[3]. Rough qualitative constraint a new concept that was proposed in this paper, it mainly using the advantages of the rough set attribute reduction to reduce the results of qualitative reasoning and improve the binding efficiency. III. THE CONSTRAINTS DESIGN OF ROUGH QUALITATIVE SOFTWARE A. The general idea of software design Software Design: Qualitative reasoning results are connect to the database, according to rough set information table in the database fill in the data table , according to rough set attribute reduction algorithm, the data tables for all items in each category reduction, to identify important items which shows in the results window. The 2nd International Conference on Computer Application and System Modeling (2012) Published by Atlantis Press, Paris, France.