Context: Software requirements must be classified for later use of requirements in the design and implementation phases. This classification can be done manually which takes a lot of time and effort or recently it can be done automatically. Automatic requirement classification is an important and promising area in both industry and research.
Objective: The objective of this paper is to investigate the existing automatic classification techniques in order to help researchers and software engineers to choose the appropriate requirement classification technique.
Method: This research was conducted with three research questions and followed Systematic Literature Review (SLR) guidelines to analyze, evaluate, interpret the relevant researches.
Results: 14 primary studies were conducted from the search processes. 2 out of them were published in Elsevier, 3 were published in Springer, 2 were published in IEEE, and the rest of them were published in different conferences and journals
Conclusions: automatic classification of SW requirements is an important task; however, it has a little work in this field. This SLR summarize all existing techniques available in this topic. It needs more effort and research papers to enhance and increase the performance as well as the quality of proposed tools.