[Paper Review] Using Data Warehouse to Support Building Strategy or Forecast Business Tend

Phuc Van Nguyen|arXiv (Cornell University)|Dec 14, 2011

Big Data and Business Intelligence2 references3 citations

TL;DR

This paper proposes a data warehouse architecture to support strategic business decision-making and forecasting by integrating heterogeneous data sources into a centralized, analyzable repository. By enabling multi-tier data marts and lifetime customer interaction tracking, the system enhances strategic planning and trend prediction through unified, historical data analysis.

ABSTRACT

The data warehousing is becoming increasingly important in terms of strategic decision making through their capacity to integrate heterogeneous data from multiple information sources in a common storage space, for querying and analysis. So it can evolve into a multi-tier structure where parts of the organization take information from the main data warehouse into their own systems. These may include analysis databases or dependent data marts. As the data warehouse evolves and the organization gets better at capturing information on all interactions with the customer. Data warehouse can track customer interactions over the whole of the customer's lifetime.

Motivation & Objective

• To develop a data warehouse framework that consolidates heterogeneous data from multiple sources for strategic business use.
• To enable organizations to forecast business trends by analyzing historical and real-time customer interaction data.
• To support multi-tier data architecture where departments can build dependent data marts from the central warehouse.
• To improve decision-making by providing a unified view of customer behavior across their entire lifecycle.
• To enhance data integration and analytical capability for long-term strategic planning in dynamic business environments.

Proposed method

• Designing a multi-tier data warehouse architecture to support enterprise-wide data integration and departmental data marts.
• Integrating data from diverse sources into a centralized data store using ETL (Extract, Transform, Load) processes.
• Implementing dimensional modeling to organize data for efficient OLAP (Online Analytical Processing) queries.
• Tracking customer interactions over time using slowly changing dimensions to maintain historical accuracy.
• Enabling query and analysis across the entire customer lifecycle to support strategic insights and forecasting.
• Utilizing the data warehouse as a foundation for building analytical databases and dependent data marts at the departmental level.

Experimental results

Research questions

• RQ1 How can a data warehouse architecture effectively integrate heterogeneous data sources to support business strategy?
• RQ2 What role does centralized data storage play in improving business trend forecasting accuracy?
• RQ3 How can multi-tier data marts be derived from a central data warehouse to support department-specific decision-making?
• RQ4 In what ways can lifetime customer interaction data enhance strategic planning and forecasting?
• RQ5 What architectural components are essential for enabling scalable and reusable business intelligence systems?

Key findings

• The proposed data warehouse architecture successfully integrates data from multiple heterogeneous sources into a unified, analyzable format.
• The system enables long-term tracking of customer interactions across their entire lifecycle, supporting deeper strategic insights.
• Multi-tier data marts derived from the central warehouse allow departments to perform targeted analysis while maintaining data consistency.
• The architecture enhances the organization's ability to forecast business trends through centralized, historical data analysis.
• The integration of OLAP and dimensional modeling supports efficient querying and reporting for strategic decision-making.
• The framework demonstrates scalability and adaptability for evolving business intelligence needs in dynamic environments.