Abstract
A business intelligence architecture comprises of different unique components to collect, transform, analyze and present the structured and unstructured raw data in simple formats to assist decision makers in making timely decisions. The introduction of service-oriented architecture (SOA) enables reusable services which are accessible over a network on demand. However, there is still a lack of academic literatures on the business intelligence architecture with service-oriented concept. Based upon various references on BI architectures from major vendors, a novel BI architecture that is built on service-oriented concept is presented and described in this paper. The proposed service-oriented architecture enables enterprises to deploy a more agile, flexible, cheaper, reusable, reliable and responsive BI applications in supporting decision making process.
Introduction
In today's fast-paced business environment, the ability to derive actionable insights from vast amounts of data is a competitive necessity. Business Intelligence (BI) systems are designed to facilitate this by transforming raw data into meaningful information. However, traditional BI architectures often suffer from being "tightly coupled." In these legacy models, the client-side presentation layer is deeply intertwined with the server-side logic. Consequently, any modification to one component often triggers a cascade of required changes across the entire system, leading to high maintenance costs and reduced agility.
Methodology: Comparative Analysis of BI Frameworks
To address these challenges, this research adopts a comparative analysis approach. The authors synthesized existing BI architectural models from industry leaders such as IBM and Microsoft, alongside academic frameworks like the six-layer model proposed by Davenport et al. (2007). By identifying commonalities and gaps in these established frameworks, the researchers sought to create a unified model that leverages Service-Oriented Architecture (SOA) principles.
The goal was to move beyond tool-specific implementations and define a robust, conceptual framework that prioritizes functional independence and interoperability. This comparative study highlights how different vendors categorize data access, analysis, and storage, providing the necessary context for a more flexible, service-oriented synthesis.
The Proposed Six-Tier SOA-Based Model
The core contribution of this paper is a novel six-tier conceptual model for BI. Unlike traditional models, this architecture introduces a dedicated "Communication Tier" that serves as the system's backbone, facilitating interaction between disparate services using open web standards like SOAP, XML, and WSDL.
The six tiers are defined as follows:
- Data Source Tier: Encompasses all raw data origins, including legacy systems, external databases, flat files, ERP, CRM, and SCM systems.
- Data Integration Tier: Focuses on ETL (Extract, Transform, Load) services, including data cleansing, standardization, and enrichment.
- Processed Repository Tier: The storage layer containing data warehouses, data marts, and operational data stores.
- Analytic Tier: The "intelligence" layer, providing services for enterprise reporting, data mining, business modeling, and OLAP.
- Communication Tier: The SOA-driven "BI Bus" that ensures location transparency and platform independence through standardized protocols.
- Presentation Tier: The user-facing layer, delivering insights via dashboards, reports, and charts across various devices like PCs, tablets, and mobile phones.
Key Insights: Decoupling and Security
A significant innovation in this model is the strategic placement of security measures. Rather than a single perimeter defense, the authors propose "Security Tiers" or firewalls between specific layers. As shown in the architectural diagram, firewalls, routers, and encryption services are placed between the Data Source and Integration tiers to protect the system from external data threats, and between the Communication and Presentation tiers to ensure that only authorized, "clean" data reaches the end-user.
Furthermore, the architecture emphasizes "Architecture Before Tools." Many BI projects fail because organizations select software before defining their structural needs. This SOA-driven approach provides a template that allows enterprises to plug in new services—such as a new analytic tool or a different data source—without reengineering the entire pipeline. This "location transparency" means the presentation layer does not need to know the technical specifics of how analytic services are hosted.
Conclusion and Future Directions
The proposed SOA-driven BI architecture offers a blueprint for more agile, scalable, and cost-effective decision-support systems. By decoupling components into reusable services, organizations can achieve greater flexibility in their IT infrastructure, allowing for the incorporation of new BI elements without the complexities typically associated with traditional upgrades.
However, it is important to note that this model is currently conceptual. Future research should focus on empirical validation through case studies or pilot implementations in real-world enterprise environments. Additionally, investigating the performance implications—such as potential latency introduced by the Communication Tier and multiple security checkpoints—would provide valuable insights for practical deployment.
Summary updated