基于大数据算法的电话号码与客户编号对应关系的识别研究与实现

RECOGNITION OF CORRESPONDENCE BETWEEN PHONE NUMBER AND CUSTOMER NUMBER BASED ON BIG DATA ALGORITHM

基于95598业务,利用大数据分析挖掘方法,构建统一身份识别模型,有效识别客户来电号码与户号的对应关系。采用大数据文本挖掘技术,有效解析用电地址信息、客户姓名等内容,并计算地址相似度得分、姓名相似度得分,作为对应关系校验以及识别疑似户号的关键因子指标。针对能获取到的对应关系,构建权重划分模型,计算对应关系匹配度得分,根据分值大小,校验对应关系的可靠性。针对找不到户号对应关系的来电号码,基于文本相似度得分构建KNN模型,计算对应关系匹配度得分,依据分值大小,识别疑似户号。

On the basis of hotline 95598, we established a unified identification model by using the method of large data analysis and mining. It could effectively identify the corresponding relationship between customer call number and household number. Big data text mining technology could effectively analyze the user s address information, name and other content, and calculate the address and name similarity score. And these similarity scores were used as a key factor index to verify the corresponding relationship and identify the matching user number. Aiming at the corresponding relationship that can be obtained, we constructed the weight partition model to calculate the matching score. The reliability of the corresponding relationship was checked according to the value of the score. When the caller ID could not be found, the KNN model was constructed based on the text similarity score, and the matching degree score of the corresponding relationship was calculated. According to the score, the matching number could be identified.