Abnormal purchase behavior detection method and system based on static information

• Main Authors: CHEN JIAQIAN, WU SHAOHUA, WU JIANGHUANG, CHEN JUNXING, SONG ZHENGCHEN
• Format: Patent
• Language: Chinese; English
• Published: 03.05.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention discloses an abnormal purchase behavior detection method and system based on static information. The method comprises the steps that the static information of retailers is acquired from a tobacco institution; performing data analysis and processing on the static information to obtain basic features, including feature creation, missing value filling and normalization processing; marking retailers according to historical case data to obtain sample data, enhancing the sample data, performing text coding model training to output text features, and inputting features obtained by fusing basic features and the text features into a classifier for model training; and carrying out risk level evaluation on the retailers by utilizing the trained model. According to the detection method, the cost of on-site inspection is greatly reduced, and meanwhile, the accuracy of anomaly recognition is also improved. 公开了一种基于静态信息的异常进货行为检测方法和系统，包括从烟草机构获取零售户的静态信息；对静态信息进行数据分析、处理，获得基础特征，包括创建特征、缺失值填充和归一化处理；根据历史案件数据对零售户进行标注获取样