Performance And Result Analysis Of Hybrid Adaptive Random Partition Testing For Improved Bug Detection And Test Coverage

Testing is a very crucial process of assuring products reliability and trust dependencies. Thus for achieving maximum coverage the test must be generated from overall distributed regions of input domains and known as partition testing. But it gives optimal results for homogeneous regions. Random testing serve better than partition testing but is also generating high computation overheads. Another technique is Adaptive Random technique having adaptive nature of regenerating and passing some of the test cases back to the input for correcting the next test cases lined to be passed. Apart from the above benefits of ART the complete problem solution of PT and RT is not provided. Thus this work proposes a Hybrid Adaptive Random Partition Testing (H-ARPT). The effective test cases can be determined if the test comes from complete regions and covers at least once each type of input. But all of certain such heavy numbers of inputs are not tested with some minimum attempts. Hence, a new mechanism is required which reduces the test size but increases the code coverage. It works towards assuring the reliability of the system. This paper proposed a Hybrid ARPT, which works towards effective and early identification of bugs according even with their priority levels also. Means the module which is most critical should be tested more. It overcomes the existing issues of high testing cost and computation complexity. The paper also suggests some analytical evaluation factors and compares the RT, PT, ART with hybrid ARPT. On the preliminary evaluations the works seems to provide effective solution of testing domains. KeywordsSoftware Testing, Blackbox Testing, Adaptive, Random and Partition Testing, Hybrid Adaptive Random Partition Testing, Coverage analysis and Fault Detection Rate;