Empirical Validation of Requirements Traceability Metrics for Requirements Model of Data Warehouse using SVM
Tanu Singh, Manoj Kumar · 2020
Data warehouse stores historical information, which is used by managers to take organizational decisions. Thus, information quality of data warehouse becomes important and assessed by the quality of its data model. Various authors proposed different metrics for assessment of the data model quality at physical, logical and conceptual level. However, very less research proposals are seen in the literature to assess the quality of requirements data model. Requirements completeness and traceability metrics for requirements model are proposed and formally validated, but no empirical validation is witnessed in the past. In this paper, empirical validation of requirements traceability metrics is performed for predicting the understandability of data warehouse requirements schemas using support vector machine. The results showed high accuracy and precision for predicting understandability of requirements schemas. In this way, quality of requirements model may be improved and subsequently used for obtaining good quality of conceptual data model for data warehouse.