民航收益漏洞NOSHOW规则提取

Civil aviation revenue leakage NOSHOW rules extraction

为减轻旅客订座后NOSHOW给航空公司造成的巨额经济损失,对海量数据的处理技术进行研究,利用粗糙集理论知识设计MapReduce算法,在Hadoop平台上对特征信息与收益漏洞NOSHOW进行关联性分析,剔除弱相关的特征维度,对约简后的特征信息使用Apriori算法进行关联规则挖掘。实验取得了不同阈值下的NOSHOW规则,利用阈值大小可以制定相应特征旅客群体的出票时限,清理不合理占座,增加座位的销售机会。

To reduce huge economic losses for airline caused by NOSHOW passengers, massive data processing technologies were studied, MapReduce algorithm based on rough set theory was used, the relationship between feature information and NOSHOW was analysed on Hadoop platform, and the weak dimensions were eliminated. Apriori algorithm was used to mine association rules. NOSHOW rules under different thresholds were obtained through experiments, and the thresholds could limit the ticket firming time for different characteristic groups, hence cleaning up the unreasonable seats, increasing the seat sales opportunities.