Абстрактный

An Efficient Algorithm for Mining Frequent Items in Data Streams

Dr. S. Vijayarani, Ms. P. Sathya

Data stream is a continuous, real time, ordered sequence of items. In data stream, data arrives endlessly and the volume of data can be potentially infinite. In recent years advances in hardware and software technologies have resulted in automated storage of data from a variety of process. Data mining techniques are applied in data streams to find out the significant knowledge. The term data mining refers, to find relevant and useful information from large database. Some of the important techniques in data mining are association rule, classification, clustering, frequent episodes, and deviation detection. Frequent pattern mining is used to find important frequent patterns from the large dataset. Click stream analysis, market basket analysis, web link enquiry, genome study, network monitoring and medicine designing are some of the important areas where frequent pattern mining is used. Most commonly used frequent pattern mining algorithms are Apriori, partition algorithm, pincer- search algorithm, fp-growth algorithm, dynamic item set counting algorithm and so on. In this research paper Éclat and Rapid Association Rule mining algorithm are used for finding the frequent item sets in data streams. The experimental results show that the performance of RARM algorithm is better than Éclat.