在資料串流的環境下探勘高效益項目集

Mining High Utility Itemsets in a Data Stream

探勘高效益項目集同時考慮每筆交易中各項目被購買的數量以及項目集可能帶給公司的利潤，以求出最有價值的商品組合。然而，交易資料會隨時間不斷改變。當資料新增或刪除時，高效益項目集也會有所改變。目前在資料串流的環境下探勘高效益項目集的演算法都需要花費很多時間再次掃描原始資料庫並搜尋候選高效益項目集，才能重新找出所有的高效益項目集，沒有利用到之前已找出的高效益項目集。因此，本篇論文提出一個在資料串流的環境下有效率的探勘高效益項目集的演算法。當資料新增或刪除時，我們的演算法不需要再掃描資料庫，也不需要搜尋候選高效益項目集，就可直接找出所有的高效益項目集。實驗結果也顯示我們的方法比之前的演算法更有效率。 Mining high utility itemsets considers the purchased quantities and the profits of the itemsets in the transactions, which can find the profible products. In addition, the transactions will be continuouslly changed over time. When the transactions are added or deleted, the original high utility itemsets will be changed. The previous proposed algorithms for mining high utility itemsets over data streams need to rescan the original database and search for the candidate high utility itemsets without using the previous high utility itemsets. Therefore, in this paper, we propose an algorithm for efficiently mining high utility itemsets in data streams, when the transaction are added or removed, our algorithms do not need to re-scan the transaction database and search for candidate high utility itemsets. Experimental results also show that our algorithm outperforms the previous approaches.

Mining high utility itemsets considers the purchased quantities and the profits of the itemsets in the transactions, which can find the profible products. In addition, the transactions will be continuouslly changed over time. When the transactions are added or deleted, the original high utility itemsets will be changed. The previous proposed algorithms for mining high utility itemsets over data streams need to rescan the original database and search for the candidate high utility itemsets without using the previous high utility itemsets. Therefore, in this paper, we propose an algorithm for efficiently mining high utility itemsets in data streams, when the transaction are added or removed, our algorithms do not need to re-scan the transaction database and search for candidate high utility itemsets. Experimental results also show that our algorithm outperforms the previous approaches.