AN EFFICIENT ALGORITHM FOR MINING FREQUENT ITEMSETS OVER DATA STREAMS UNDER THE TIME- SENSITIVE SLIDING-WINDOW MODEL

Mining frequent itemsets has been widely studied over the last decade, mostly focuses on mining frequent itemsets from static databases. In many of the new applications, data flow through the internet or sensor networks which extend the mining techniques to a dynamic environment. The main challenges include a quick response to the continuous request, a compact summary of the data stream, and a mechanism that adapts to the limited resources. Here, we propose a time-sensitive sliding window model for mining frequent itemsets from data streams. Our approach consists of a storage structure that captures all possible frequent itemsets and a table providing approximate counts of the expired data items, whose size can be adjusted by the available storage space. This approach guarantees that both the execution time and the storage space remain small under various parameter settings. In time critical applications, a sliding window model is needed to remove old data. So, we adopt a model to mine the K most interesting itemsets, or to estimate the K most frequent itemsets of different sizes in a data stream. This method partitions the sliding window into buckets. This model provides frequency counts of the itemsets for the transactions in each bucket and supports to find out the top-k frequent itemsets. This method shows that it utilizes very small memory space