Semi-supervised classification of non-functional requirements: An empirical analysis

AGUSTÍN CASAMAYOR; DANIELA GODOY; MARCELO CAMPO

INTELIGENCIA ARTIFICIAL. IBERO-AMERICAN JOURNAL OF ARTIFICIAL INTELLIGENCE

Asociación Española para la Inteligencia Artificial (AEPIA)

Año: 2009 vol. 13 p. 35 - 45

1137-3601

The early detection and classification of non-functional requirements (NFRs) is not only a hard and time consuming process, but also crucial in the evaluation of architectural alternatives starting from initial design decisions. In this paper, we propose a recommender system based on a semi-supervised learning approach for assisting analysts in the detection and classification of NFRs from textual requirements descriptions. Classification relies on a reduced number of categorized requirements and takes advantage of the knowledge provided by uncategorized ones as well as certain properties of text. Experimental results show that the proposed recommendation approach based on semi-supervised learning outperforms previous proposals for classifying different types of requirements.