研究了一种新的增量数据挖掘方法，通过对交易数据项集进行编码后，把原始数据转换成整数值随机变量序列，得到该序列为马尔可夫链，利用频率代替转移概率，建立了一个趋势分析的模型。通过对超市销售数据的分析表明，该方法具有简单、实用的特点，得到了顾客对同一产品的不同品牌的选择是没有差别的。关 键 词 数据挖掘; 趋势分析; 转移概率矩阵; 增量Abstract In the paper, a new increment method of data mining is presented. Based on encoding the transaction data item set, the original data is transformed into a integral variable series is proved to be a Markov chain in theory. And instead of transfer probability, frequency is used in transfer probability matrix. Then a module is built for trend analysis .The analysis for the sale data of supermarket indicates that the method is fine, and a nice result that customs choice is same to different bland of one product is received.Key words data mining; trend analysis; transfer probability matrix; increment