Abstract:
"
The Software Requirements Specification (SRS) document is the most detailed and influential
document in the software development process, this information is highly rich in content and is
contained with confidential data. All subsequent steps during the process of development are
decisive by this document. The significance of the document is that it is modelled for a clear
depiction of the proposed system’s features as requirements which can be referred by the project
team to identify all tasks in relation to deliver the end product in accordance to their respective
requirements.
Testing is a validation activity that if performed in order to ensure the conformance of software
systems with respect to their functional and non-functional requirements and specifications. The
assurance of the quality of the product is supported by every requirement being tested. The failure
of a test case is often mapped with the software that was written for the requirement offering
traceability of a defect.
There is a growing attentiveness given to Machine Learning approaches for the automation of
various tasks. As such there are approaches to automate many processes in the software testing
cycle. The challenge in the automation process is it requires human intelligence to interpret the
required output. The decision to use supervised text classification techniques are to translate the
requirement that are in natural language to a logical format that can be validated and generates test
scenarios from them.
This research hereby intends to define the significance of the requirements specified in the
requirements specification document of software products and facilitate an automating the process
for the of creation of scenarios to derive test cases in the software testing cycle, this will prove to
improve end-to-end testing by identifying the right test scenarios."
