Font Size: a A A

Software Test Study Of Rough Set Based Model

Posted on:2008-04-18Degree:MasterType:Thesis
Country:ChinaCandidate:C Z WangFull Text:PDF
GTID:2178360218956636Subject:Traffic Information Engineering & Control
Abstract/Summary:PDF Full Text Request
Software Test is becoming an important factor restricting the development of software. At present the methods in software test research are based on some complex and hard understood theory, or having too complex operations to use. RS theory is a kind of math tools used to deal with uncertain, imprecise and antipathy data, it needn't any additive suppose or condition. Moreover, the rules gained by RS method are easily understood associated rules that accord with people's experience and more suitable for the application in manager decision. Therefore it is will be a new research field of great value making the use of RS theory to software test.Now the work and correlative literature about the study and application of the RS in software test and soft engineering is extraordinary rare, the research in this new domain hasn't been developed almost. The study of this paper is just based on this background and starts on the new research field.The study of this paper has two main parts which are also the innovations of this paper. The first one is the study about RS arithmetic: based on the sum-up and analysis of present RS data reduction arithmetic, the paper advances new attribute and attribute value reduction arithmetic with complexity's analysis. The later part is the study of the RS based modeling in software test: there are three RS based test models are brought forward.At the beginning, the paper introduces the background and corresponding knowledge: the uncertainty and the technology of software test; RS theory's basic concepts, data reduction methods and its characteristic thereby educes the think of the paper choosing.Secondly, considering the feasibility of coding and arithmetic complexity synthetically, new attribute reduction arithmetic and attribute value reduction arithmetic are advanced.In succession, according to the uncertainty in software test, three RS based test models are brought forward: test case optimizing model; software quality evaluating model and quality ordering model. The test case optimizing model is used to reduce the redundancy data including in the original test case, the later two models are applied to evaluate the grade of one in a kind of software, or order some congener software.Finally, the improved RS data reduction arithmetic are adopted in the design of RS based software test manage system (RS_TMS), then the kernel function module——RS data reduction module is designed and realized.The paper gives a conclusion of the study and advances a prospect of the study on the new field that the RS's application on software test, both and the software engineering.
Keywords/Search Tags:Software Test, Uncertainty, Rough Set(RS), Ordered Information Table, Model
PDF Full Text Request
Related items