Requirements Modeling and Quality Issues for Development of Data Warehouse

Abstract

As Data Warehouse (DW) is very crucial for the organizational decision making, continuous attention must newlinebe paid for evaluation of quality throughout its design and development. Data warehouse newlinequality depends on the quality of its requirements, conceptual, logical and physical newlinemodels. Some of the authors have proposed guidelines and metrics to assure the quality of newlineconceptual, logical and physical models for DW. There is no relevant work found in the newlineliterature to assure quality of requirements model for DW. Moreover, a set of structural newlinemetrics to assess quality of conceptual models for DW has been suggested in the literature newlineand significant relationship between metrics and understandability of DW conceptual newlineschemas has been found using various correlation analysis techniques such as newlineSpearman s, Pearson etc. However, advanced statistical and machine learning methods newlinehave not been used to predict effect of each metric on understandability. These also newlinemotivate us to propose in the thesis, a quality-oriented requirements engineering approach newlinefor DW, where (i) agent-goal-decision-information (AGDI) model is extended with the newlinenotion of perspective to capture intentions of the agents,(ii) based on proposed model of newline(i), DW requirements are captured through various modeling activities, (iii) in order to newlineassure quality of requirements model, various guidelines are proposed that must be newlinefollowed by the analyst while performing various modeling activities of (ii), (iv) newlinerequirements completeness and traceability metrics are proposed to monitor the progress newlineof DW requirements engineering (RE) products (i.e. requirements model). The proposed newlineguidelines and metrics have been successfully applied to the DW requirements models for newlinetypical Banking Organization.