University Links: Home Page | Site Map
Covenant University Repository

A framework for an Integrated Mining of Heterogeneous data in decision support systems

Afolabi, I. T. and Uwadia, C. O. and Ayo, C. K. (2008) A framework for an Integrated Mining of Heterogeneous data in decision support systems. In: First International Conference on Mobile-computing, Wireless Communication, E-Health, M-Health and TeleMedicine.

[img] PDF
Download (413kB)


The volume of information available on the Internet and corporate intranets continues to increase along with the corresponding increase in the data (structured and unstructured) stored by many organizations. Over the past years, data mining techniques have been used to explore large volume of data (structured) in order to discover knowledge, often in form of a decision support system. For effective decision making, there is need to discover knowledge from both structured and unstructured data for completeness and comprehensiveness. The aim of this paper is to present a framework to discover this kind of knowledge and to present a report on the work-in-progress on an on going research work. The proposed framework is composed of three basic phases: extraction and integration, data mining and finally the relevance of such a system to the business decision support system. In the first phase, both the structured and unstructured data are combined to form an XML database (combined data warehouse (CDW)). Efficiency is enhanced by clustering of unstructured data (documents) using SOM (Self Organized Maps) clustering algorithm, extracting keyphrases based on training and TF/IDF (Term Frequency/Inverse Document Frequency) by using the KEA (Keyphrases Extraction Algorithm) toolkit. In the second phase, association rule mining technique is applied to discover knowledge from the combined data warehouse. The final phase reflects the changes that such a system will bring about to the marketing decision support system. The paper also describes a developed system which evaluates the association rules mined from structured data that forms the first phase of the research work. The proposed system is expected to improve the quality of decisions, and this will be evaluated by using standard metrics for evaluating the interestingness of association rule which is based on statistical independence and correlation analysis.

Item Type: Conference or Workshop Item (Paper)
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: Faculty of Engineering, Science and Mathematics > School of Electronics and Computer Science
Depositing User: Mr Adewole Adewumi
Date Deposited: 18 Feb 2014 10:54
Last Modified: 18 Feb 2014 11:18

Actions (login required)

View Item View Item