Feature extraction and performance measure of requirement engineering (RE) document using text classification technique
L. P. Saikia, Shilpi Singh · 2018
The RE document in the SDLC phase of software development is prone to ambiguity, since it is written in natural language. The text classification is a method of assigning a document as predefined classes or categories. The efficient understanding of text document is important to improve the quality of RE document, and this can be achieved by using semantic information regarding a text document. The main objective of this experimentation is to utilize semantic information to identify features and prepare data sets for better classification of text as “Ambiguous” or “Unambiguous”. The different data sets are constructed and then are analyzed both manually as well as computationally on different parameters (kappa index and likelihood ratio) to understand the quality of any of RE documents.