Deriving Context Utility of Test Cases using Fuzzy Logic

Abstract: “Thousands of test cases executed, Hundreds of QA person hours spent, Product release opens a Pandora’s Box of Issues!”

Sounds familiar? What is the root cause? Most often than not, it is observed that misalignment with customer usage patterns is the culprit. The outcome of the testing effort depends on the utility of the test case chosen for that particular context. How does the tester choose which test cases to execute within the given constraints is a fundamental problem. The test scope must prioritize the flows with the highest context utility. These can only be derived once the entire context oriented factors and constraints are taken into consideration.

This presentation talks about Optimal Contextual Scoping where test cases are filtered through Fuzzy Logic based smart channels which are dynamic and can vary based on the context. These channels are factors which make a test case relevant or irrelevant for inclusion in the scope. This presentation addresses factors like Criticality, Complexity, Code Change, Customer viewpoint, Time, Resources and is dynamic enough to include many others. Each channel has some fuzzy values defined in the fuzzy set and based on these values optimal test cases can be derived.

Other Optimization techniques are not flexible enough to use in multiple contexts and also there is no surety if all the high context utility flows are covered. This paper examines a solution where the best set of test cases can be derived based on the context of the certification.