SIMULATION OF IMPROVED ASSOCIATION RULE DISCOVERY SYSTEM FOR DECISION SUPPORT

Macarthy Osuo-Genseleke, Asagba Prince O

Abstract: Recent computing transactions entails large sum of data which are retrieved, stored and used for operations. The data often contain association relationships which can be mined to aid management decision. We simulate a data mining system for the association of rule discovery using an improved C4.5 Algorithm. The system extracts data and its close relationships that will be used for decision making. The system analysis and its design was done using the Object - Oriented System Analysis and Design Methodology (OOADM). Weather data file was used to test run the system and results shows that mining associated rules in a large database is important in decision making. Algorithms and Designs in rule discovery associations that seem complex but very useful in making decisions need to be well implemented to be useful to users. This paper presents a simulation of an improved C4.5 Algorithm for association rule discovery system used in decision support. Java programming language is used for the implementation with Netbeans IDE. When the system was tested, users reported better usability, efficiency and clarity of results from the application.

Keywords: Rule Discovery, Data Mining, C4.5, Decision Tree,Algorithm, Association Rule.

DOI: https://doi.org/10.15623/ijret.2018.0706018