Conceptual Based Hidden Data Analytics and Reduction Method for System Interface Enhancement Through Handheld devices

Abstract

With the increasing demand placed on online systems by users, many organizations and companies are seeking to enhance their online interfaces to facilitate the search process on their hidden databases. Usually, users issue queries to a hidden database by using the search template provided by the system. In this thesis, a new approach based mainly on hidden database reduction preserving functional dependencies is developed for enhancing the online system interface through a small screen device. The developed approach is applied to online market systems like eBay. Offline hidden data analysis is used to discover attributes and their domains and different functional dependencies. In this thesis, a comparative study between several methods for mining functional dependencies shows the advantage of conceptual methods for data reduction. In addition, by using online consecutive reductions on search results, we adopted a method of displaying results in order of decreasing relevance. The validation of the proposed designed and developed methods prove their generality and suitability for system interfacing through continuous data reductions.