Predictive Analytics and Software Defect Severity: A Systematic Review and Future Directions

Table 4

Data extraction from.

S/N	Dimension	Attribute: description
1	The type of severity prediction class adopted	Probability, severity levels, or binary classification
2	Data sampling status and metrics characterization	The distinction between balanced or imbalanced data together with data attribute type either of nominal or numeric
3	Category of machine learning used and learner algorithm used for prediction	Choice of either supervised or unsupervised with learner algorithm used for prediction
4	Parameter tuning or optimization strategy adopted	Identification of performance improvement tuning adopted in literature
5	Threats to validity with respect to future work suggestion	Mapping of primary studies threats to proposed future work outlook in the software defect prediction industry
6	Threat to validity and future work recommendations	Juxtaposing vulnerability with future potentials
7	Use case application area	Identifying application area of either cross-project or within the project use case