University of Twente Student Theses

Login

Opportunities and limitations of using SOA concepts and technologies for building BI applications : a Delphi study

Linders, Stefan (2008) Opportunities and limitations of using SOA concepts and technologies for building BI applications : a Delphi study.

[img] PDF
1MB
Abstract:Service-Oriented Architecture (SOA) is gaining increasing attention to improve the flexibility and agility of organizations. However, the current focus is mainly on supporting operational systems, not on supporting Business Intelligence (BI) systems. This thesis aims to identify potential benefits or challenges of using SOA for the organization and use of BI. The thesis is guided by the following research question: “What are the opportunities and limitations of using SOA concepts and technologies for building BI applications?” To answer this question, we first performed a literature study on the concepts and technologies of both BI and SOA. After that, we applied a Delphi study to a panel of experts in order to generate and discuss opportunities and limitations. In this study, we used models from our literature study for guidance of the experts. The result of this Delphi study is a ranked list of opportunities and limitations, which include the arguments and comments from the experts. We have further investigated a small subset of these opportunities and limitations, by using the comments from the expert panel, having private interviews with some of our experts, and consulting additional literature. Literature study on SOA and BI We identified BI as the process of gathering and analyzing data, and using the produced information to steer the organization. The process consists of four phases, (1) planning and direction, (2) collection of data, (3) analysis of data, and (4) distribution of data. We also identified the areas of the organization for which BI delivers information, and what this information is about. We defined SOA as an application architecture within which all applications logic is defined as services, which can be called in defined sequences to form business processes. We discussed how services interact and to what principles they should adhere. The main benefits that would justify the use of SOA are improved possibilities for both reuse and integration, which could result in increased agility and adaptability of the organization as a whole. Opportunities The majority of the opportunities that have been identified and rated correspond to the reuse and integration benefits that are also associated with SOA. Our main observations on the identified opportunities are as follows: Reuse of services for the collection of data and distribution of information. Integration with operational systems for the collection of data. SOA is perceived as offering capabilities for better integration of (operational) systems with the BI systems. Integration with operational systems for the distribution of information, to improve the use of BI information in operational processes. Integration of the components of BI systems through the use of services could result in a more flexible BI architecture. For the tactical and operational level more opportunities are identified than for the strategic level, probably because BI is most often used at these levels, and because they have a greater need for a flexible BI organization that can adapt to the more frequent changes at these levels. No considerable differences in the value of opportunities for different focus areas have been identified. We have further investigated Business Activity Monitoring (BAM), which is one of the identified opportunities. BAM supports the steering of operational processes by providing real-time information on the current state of these processes. BAM systems need to collect data from various operational systems, and SOA provides the means for integrating these systems. Part of this support is found in event-based messaging, which enables real-time data collection. We also further investigated Master Data Management (MDM). MDM serves to maintain a consistent definition of business entities throughout operational systems, and can also store the data of those entities in a central location. MDM can provide consistent and up-to-date data on entities to the BI system, and can improve the consistency of the data analyses. SOA is identified as a suitable delivery system to integrate the operational systems with the MDM system. Furthermore, transformation and cleansing services could be used to build the MDM system. Limitations The most serious and most discussed limitation in the Delphi study is the transportation of large data sets over web services. Large sets of data often need to be transported from the operational systems to the data warehouse, for the collection of data for BI. Web services, often employed for the exchange of rather small messages, currently do not seem suitable for transporting large data sets. Although event-based communication is regarded a possibility for using services for the collection of data, we have no knowledge of viable implementations of this concept. Further research on handling large data sets in an SOA is needed to find viable approaches for realizing actual benefits of handling this data for the collection of data for BI. Conclusion Overall we regard the approach of this research a success. Many opportunities and limitations of using SOA concepts and technologies for building BI-applications have been identified, and some of them indicate several benefits for organizations. This research has touched on many subjects concerning SOA and BI that are only abstractly mentioned, and therefore provides a starting point for further research into those subjects.
Item Type:Essay (Bachelor)
Clients:
Capgemini
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Business & IT BSc (56066)
Link to this item:https://purl.utwente.nl/essays/58525
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page