数据驱动的证券客户画像及流失预测分析

Data-driven Securities Customer Portrait and Customer Loss Prediction Analysis

客户流失是当前证券公司面临的严峻挑战。客户画像能精准掌握客户动态和意向,动态计算客户标签并及时提供公司所需的信息,极大地提高了数据的价值和智能应用。从证券客户的异构多源数据集出发,给出证券客户画像及流失预测框架,包括多源数据、数据治理、客户画像和流失预测4个模块。基于大数据技术深入挖掘客户的证券业务数据和行为数据,对客户的兴趣、性格、职业、信仰等个性特征做出准确描述,并采用算法实现对证券客户的特征选择和画像。进而,融合大数据和人工智能技术,提出客户流失预测模型和实现步骤,为提高证券客户数据管理效率、服务能力提供相应的方法。

Customer portraits can accurately grasp customer dynamics and intentions,dynamically calculate customer targets and provide timely information required by the company,which greatly improves the value of data and intelligent applications.Based on the construction of digital resources for securities customers,the portraits of securities customers and the customer loss prediction framework are given,and the functions at different levels are analyzed.The portrait contains comprehensively using the user’s securities business data and behavioral data,accurately describing the user’s personal characteristics such as interest,personality,occupation,and belief,and using algorithms to realize the feature selection and portrait of securities customers.Furthermore,it integrates big data and artificial intelligence technologies to propose a customer loss prediction model and analysis steps,and provides the corresponding methods for improving the efficiency of securities customer data management and service capabilities.