A Novel Information Lifecycle Management Scheme for a Web-Based Database System

In this paper, we provide details regarding the design, analysis and experiments of a new information lifecycle management scheme for a web-based database system referred to as Adaptive Information Lifecycle Management (AILM). AILM is a novel scheme because it simultaneously considers variations in time and data usage. In addition, AILM consists of four main components: data collection, information valuation, information classification, and growth prediction. In this paper, 2255 teaching cases over eight years are managed using the AILM scheme from the life.edu.tw web site. Using AILM, the growth in storage capacity increases by almost 6 similar to 7% per year, and the number of teaching cases is projected to increase by almost 160 per year in the next five years. To enhance the performance of the web-based database system, we further classify the data into three groups, namely, Class_1, Class_2 and Class_3. The experiments demonstrate that Class_1 includes approximately 50% of the teaching cases but only uses approximately 40% of the storage capacity. In addition, the hit rate has increased to 75%. Moreover, we can delete the teaching cases with a 2.5% hit rate in Class_3, which contributes to saving approximately 10% of the storage capacity. In other words, the AILM scheme is useful because it not only expedites search time but also improves storage utilization in a web-based database system.