DEVELOPMENT OF A HYBRID MODEL FOR ENHANCED BUSINESS INTELLIGENCE PROCESS

SOURCE:

Faculty: Physical Sciences
Department: Computer Science

CONTRIBUTORS:

Ngozi F. Efozia
Anigbogu S. O.

ABSTRACT:

Majority of enterprises of various industries rarely utilize the vast internally and externally generated data available in the business for productive and improved performance of the sectors. This has been due to the paucity of reliable intelligent systems to enable such sectors to fully manipulate these vast heterogeneous data. The concept of Business Intelligence comes handy to resolve the problem involved in decision making as well as in taking actions based on facts and evidence-based knowledge insight particularly in the health sector. Being an intelligent tool, it needed an effective and enhanced data integration technique to handle the vast heterogeneous data across the nation’s health centres, hence this research. The aim of the study was to develop a hybrid model for enhanced Business Intelligence process (HMEBIP). It combined ontology-based and virtual data integration (OBDI and VDI) techniques to enhance the data integration process in Business Intelligence (BI) environment. The objectives were to: provide a database for storing and tracking disease outbreak and control for resolving its historical data in real-time; provide an integrated patient medical record for accessibility across health centres in the country; use the combined data integration techniques and case-based reasoning (CBR) to boost the intelligence expertise of the Business Intelligence process as well as for resolving latencies, redundancy, and interoperability issues; and to provide the medical experts with the facilities for making accurate, intelligent, fact and evidence-based knowledge insight decisions on patients’ matters. The hybrid model was developed with Java Script, Hypertext Pre-Processor (PhP), and My Structured Query Language (My-Sql) programming languages using object-oriented analysis and design methodology (OOADM) and the model was tested with the health sector domain that deals with disease control procedure. The model was verified and validated using confusion matrix which was used to carry out the performance evaluation by comparing the result attained when only either of the two data integration techniques; ontology-based and virtual data integration (OBDI and VDI – data virtualization) was exclusively used to enhance a Business Intelligence (BI) process. The results obtained show that the hybrid model had a higher enhanced Business Intelligence (BI) process performance of 95% as against 75% and 65% for the OBDI and VDI respectively. This shows that the hybrid model is successful as it outperformed the existing model with 20% performance level.