Early determination of Optimal Test Flows with Requirements Analytics
Mukundan Sundararajan, Priti Srikrishnan, Kiran Nayak, Siddharth K Saraya · 2019
Optimal test flows in software testing involves a nuanced approach of combining knowledge of multiple test scenarios that traverse common paths or test scenarios that are segments of one or several paths based on the optimizing cost function. Identification of these flows and obtaining related information - optimum number of test cases, dependencies, complexity of requirements early in the development lifecycle is a key to realistic planning, dependency management and optimum test strategy in project delivery. An early stage view - from a working set of requirements - into the optimized flows and their complexity enables determination of resource needs, and dependencies. The paper proposes a method to obtain optimal test flows from requirements. Natural language processing of requirements enables determination of attributes, rules and relationship amongst the requirements that provides information on the number of flows, minimum necessary yet adequate test flows to ensure coverage and expose severe defects. It also provides information of the likely dependencies. Project results from application of the method support the proposed computation model. Applying this method enables early determination of optimal test flows, dependencies to identify an apt test window and key flows to uncover high severity defects that minimizes risks to project's planned schedule and product quality.