A data warehouse stores historical information, which managers rely on to make organizational decisions. Since a lot of administrative decisions are based on the data warehouse information, it is crucial to keep a constant focus on quality when designing and developing a data warehouse. To evaluate the physical, logical, and conceptual aspects of data model quality, several researchers have put forth several metrics. However, there has been limited focus in the literature on the quality assessment of metrics for requirements data models. To properly assess the metrics efficacy, empirical validation is required. These metrics must be empirically validated to accurately evaluate their effectiveness. This paper firstly implements the k-means clustering techniques on the models understandability and divide them into two clusters, i.e., understandable or non-understandable. Further, the prediction of requirements model understandability was performed by implementing the decision tree technique to assess the traceability metrics. The results demonstrate high accuracy in predicting the understandability of requirement schemas, thereby enhancing the quality of requirements models, which can ultimately lead to improved multidimensional models for data warehouses.