农业垂直搜索引擎中的价格数据来源于各个农业网站,由于多种因素,采集到的数据中存在大量异常数据。同时,采集到的农产品价格数据具有其自身的特点,这些特点使得传统的异常数据检测方法不能够很好的工作。针对搜索引擎采集到的海量农产品价格数据,提出了一种异常价格数据的检测方法。实际应用表明,该方法能够取得很好的效果。
Agriculture price data from vertical search engines are collected from various agricultural sites. Due to various factors,there are a large number of abnormal data. Since the characteristics of these agriculture price data make traditional methods of detecting abnormal data not work well,this paper presents a new method of detecting abnormal data. Practical applications show that this method can achieve good results.