Preference Computation : Skyline Query, Data Mining and Knowledge Discovery

Abstract

Abstract—We introduce a novel approach for a solution to skyline query problems that are usually detected in any online and direct marketing. A preference computation is proposed to handle a complexity of high-dimensional case of knowledge discovery that contain in the online transaction. Our reviving algorithm evidently succeeded to identify preferable objects with the lowest additional cost in the query. A concise real world example is provided to support the computational example while emphasizing the importance of knowledge discovery process in the solution. We extend the capability of a skyline query to solve the market positioning evaluation of the dominated points of products. This capability can assist both manufacturer and customer to plan product features according to their approximate distances to the preference points. For this purpose we develop a distance measurement on a convex skyline approach. First, we present data sets contain record of multidimensional product, where every dimension represents one attribute of product feature. Then we evaluate the skyline query of a data set and divide the data set into a collection of preferable objects in skyline and another are the dominated points. Here we assume that each dominated point is potentially entering the preferable region by moving their attribute values into customer preference’s points. We provide the query to find potential products to enter the skyline with a lower additional distance (as cost). This approach compute minimum additional cost to revive the dominated points based on a user’s elicitation of a maximum threshold. Results of our comprehensive experiments show the effectiveness of this approach both in real world and synthetic data sets.