EBusiness analytics framework (EBAF) - to enable SMEs to gain business intelligence for competitive advantage

Recent technological advances have resulted in increasingly larger databases. The fast and efficient useful analysis and interpretation of this data to improve business intelligence is critical to the success of all organisations. This thesis presents a new framework that utilises a new multilayer mining theory and is based on business intelligence methods, data mining techniques, online analytical processing (OLAP) and online transactional processing (OLTP). Existing decision making modelling approaches for executive information systems have three main shortcomings and limitations to different degrees: a) problems in accessing new types and new structures of data sources; b) failing to provide organizational insight and panorama; and c) generating excessive amount of trivial information. The hypothesis of this research is that a new proactive Multidimensional Multilayer Mining Management Model (5M) framework which is proposed in this thesis will overcome the shortcomings listed above. The 5M framework is made up of 6 components: (a) multilayer mining structures; (b) measurable objectives conversion models; (c) operational transaction databases; d) object-model data marts; (e) data cubes and (f) core analysis engine which analyse the multidimensional cubes, multilayer mining structures and the enterprise key performance indicators. The 5M framework was evaluated by developing an implementation of an instance of the framework called the Ebusiness Analytical Framework (EBAF). The 5M framework and the subsequent EBAF framework were built by carrying out action research and a case study in an ebusiness company where it was subject to implementation, reflection, adaptation and improvement in order to fulfil the requirements of the hypothesis and that of a real business. EBAF implemented all 6 components of the 5M framework using the Visual Studio environment and using various algorithms, tools and programming languages. The programming languages used included MDX, DMX, SQL, VB.Net and C#. Further empirical case studies can be carried out to evaluate the effectiveness and efficiency of the 5M framework.