A Study on Database Systems for Multi-Dimensional Web-Structure Mining

点击翻译按钮获取中文摘要

Web-Community Mining is a significant issue in today's cyberspace technology. One problem is how to find outstanding communities in a big data space, and the other is how to support user-given personalization in the web-community mining function. This research is aimed at solving the latter issue. The proposed solution is a new database system which provides a data-cube query model on a target web-space dataset. Namely, under a given multi-dimensional constraint, the system computes web-community structures under the constraint and returns a ranked list of communities. The proposed data-cube model has three dimensional constraints of FROM-type, TO-type, and the time of a data snapshot. Efficient set-oriented data operations on the data-cube model and their algorithms are also proposed.