Research on methods of improving customer profile in electric power marketing based on big data analysis of customer's electricity address
Hong‐Wen Lin, Yan Chen, Jiakui Zhao, Zhongping Xu, Yuze Chen, Jian Liu, Hong Wei Ouyang, Bao Yuan, Genxin Xiong · 2017
The paper is aimed at solving the problems of inconsistency, inaccuracy and non-real time of customer profile in the electric power marketing information system, and proposes the new methods of improving the customer profile in electric power marketing based on the big data analysis of customer's addresses. The core idea of the methods is that by the introduction of other industry data resources, the electricity address provided by electric power marketing information system and the customer address provided by other resources are analyzed with the big data technologies to achieve precise matching of these addresses. The address matching is followed by three steps: Chinese word segmentation with Hidden Markov Model algorithm, sentence similarity calculation based on the semantic and word order, and finally geographical coordinates matching for cross validation. Based on the matching results, the procedures of updating customer profile are processed respectively according to customer contact changes or changes in customer property rights, to achieve the consistency, accuracy and real-time data of customer profile. By implementing the automatic and intelligent operations of perfecting customer profile, the strong information support is provided for the marketing departments to enhance the “Internet+” electric power marketing services.