Reliability Knowledge-based Software Testing Model And Its Application

With the rapid development of the new generation of information technology,the software has been widely used in many fields.Software quality becomes a key concern that has a direct influence on efficacy of information technology.Software testing is an important means to guarantee the software quality.It has been one of the focuses of software engineering technology.As an important metrics of software quality,software reliability is a hot topic in the research field.The Aim of software reliability testing is to assure and validate the software reliability.However,there are many problems in software reliability testing due to the increasing complexity of software.First of all,knowledge and experience are generally beyond the scope of ordinary testers,which leads to unsatisfied testing results.Secondly,the architecture and underlying feature elements of software system are intricate,the software running and transition states increase exponentially,which leads software reliability testing difficult to be completed within a limited time and deployed in engineering applications.Some studies proposed that the knowledge based software testing method can improve the testing quality.Nevertheless,a systematic and complete set of software testing techniques based on reliability knowledge is necessary to promote the software quality and reliability.These techniques involve how to improve and optimize the software testing process,and introduce software reliability knowledge to achieve high-quality software testing in engineering practice.This dissertation devotes to solving the following questions: 1)how to introduce and use knowledge in the testing process,2)how to measure the amount of knowledge required for software testing,3)how to describe the relationships of knowledge,software test quality and software reliability,4)how to evaluate the growth of test quality after the introduction of knowledge.This dissertation introduces the situations of the software reliability testing techniques as well as their models at home and abroad.It focuses on the software testing process and method based on reliability knowledge,proposes Reliability Knowledge-based Software Testing(RKST)model,and studies measurement methods of knowledge and software testing quality.Then,the Reliability Knowledge-based Software Ontology(RKSO)base is constructed and the effectiveness of the proposed methods is verified by a case study on domestic Linux operating systems.The main contributions of this dissertation are listed below:(1)RKST model and its implementation process.Combined with software testing process and testing activities based on software life cycle,the engineering-oriented reliability knowledge ontology base is introduced to achieve the optimizations of software testing strategy and process.Besides,the RKST model is constructed,which involves software testing process and defines the cyclic learning process.It focuses on the problem of how to introduce and utilize reliability knowledge in the testing process,and provides guidance for the implementation of software testing in engineering.(2)The information entropy-based knowledge qualification and testing level measurement methods.Through the introduction of information entropy theory,this dissertation defines various types of knowledge measurements.Moreover,it can be concluded from the Vila Diagram that the larger the mutual information of testers,the better results of software testing.By the calculation of the information entropy of problem sets in software testing,this dissertation also proposes the concept of testing quality entropy and uses it to measure software testing quality.(3)The inversely proportional relationship between the scale parameter of Weibull distribution and knowledge quantity.Accordingly,with the help of the distribution curve of knowledge quantity and defect discovery,the trend of finding problems along with software reliability can be predicted,and then the end time of test can be determined.(4)Designs the architecture of RKSO base,which consists of three parts,namely the domain knowledge ontology base,reliability knowledge ontology base and reliability rules.The effectiveness of RKSO base in engineering practice is demonstrated via the testing applications and construction of domestic operating system RKSO base.The study of reliability knowledge-based software testing techniques is a comprehensive research issue involving software testing theory,knowledge ontology theory,software reliability theory,and engineering applications.An independent can perform reliability knowledge-based software testing with RKCTST model.It solves the problems of software reliability testing and reliability assurance of complex software in engineering,which enriches the methods of software testing and testing quality assessment and improves the testing efficiency.Therefore,from the point of view of practical significance,the RKST model can provide concrete practical guidance in software testing activities,and it is also useful as a reference for internal testing in research and development institutions,especially for the testing of large scale complex software systems.In addition,the reliability knowledge based testing method of RKST model can provide theoretical guidance for the development of new software testing tools.Last but not least,from the point of view of significance of academic research,it is expected that the research results of this dissertation can promote the development of relevant professional disciplines,such as software testing and software quality engineering.