关键词: 数据流, 数据流挖掘, 频繁项

Abstract: Because of the rapid data arriving speed and huge size of data set in stream model，it is usually unable to find all the accurate frequent items of a data stream.The space complexity and the time complexity are the main measurement which is used to evaluate the strongpoints and weaknesses of algorithm.This paper proposes an improved algorithm based on principle of locality to find ε-approximate frequent items of a data stream，its space complexity is O（1/ε）.The processing time for each item is O（1/ε） in the worst and the processing time for each item is O（1） in the best.Moreover，the frequency error bound of the results returned by the proposed algorithm is ∑_(i=2)^j（1-μ_i）×k_i.

Key words: data stream, data stream mining, frequent item