Implementation of AFS approach for web data analysis

Abstract

Axiomatic Fuzzy Set (AFS) theory was proposed by Professor Liu Xiaodong [1] to establish a totally new system of fuzzy sets and create a flexible methodology for the development of intelligent systems. The AFS theory provides an efficient framework that bridges real world problems with abstract constructs of mathematics of fuzzy sets. In this project, the basic AFS concepts are studied. In particular, how AFS simulates a simple real world problem is elaborated. The main work of this project concerns the transformation of AFS functions from MatLab code into Java. This is to explore the possibility of extending this technique for web data analysis. In this report, the proposed design and implementation of the AFS in Java is given. The class diagram is drawn using UML. All the 25 MatLab AFS codes are documented using flowchart, and 9 out of them have been implemented using Java codes. Last but not least, future works are suggested.