A my page service realizing method by using market expectation engine

Abstract

This paper proposes a method to realize My Page Service using market expectation engine. There are two problems on expecting market trend using My Page; (1) difficulty of analyzing huge database that manages the enormous number of customer data and the extremely broad areas that the customers might be interested in, and (2) difficulty of grasping market trend using customer data that is stored in database. We address problem (1) with threedimensional vectors that consists of customer, preference category and time axes. One of the problems of three-dimensional vectors is its huge volume of information. Our method addresses this problem by recording the positions and the values of the points only where the information has changed as time passes. And we address problem (2) with clustering customer preference data. Furthermore, we have found a few trend leaders in the groups. Using trend leaders data, we can expect market trend. © Springer-Verlag Berlin Heidelberg 2005.