宁海县个人用水用户信用评估数据

为贯彻落实信用信息基础库的建设，采集在经济活动中产生的用水数据和缴费行为数据，通过数据挖掘技术，评估用户的信用水平。通过监测信用评估数据，制定风险预警机制，及时识别可能存在的恶意欠费或用水安全隐患用户，采取措施保障供水安全。1.收集用户历史缴费数据，包括上一年度用水总量、上一年度费用、上一月份用水总量、上一年度同比用水总量、上一月份费用产生日期、开始缴费日期、逾期日期、实际缴费日期 2.根据数据分析用户行为失信特征数据（逾期次数A1、最大逾期天数A2、平均逾期天数A3）、按期缴费行为守信特征数据（按时缴费比例B1）、用水行为特征数据（上一年度平均每月用水总量、同比差异百分比C1=(上一月份用水总量-上一年度同比用水总量/12)/(上一年度同比用水总量/12)、平均每月用水差异百分比C2=(上一月份用水总量-上一年度平均每月用水总量)/上一年度平均每月用水总量） 3.将上述已归纳的用户失信特征、守信特征、用气特征进行计算,分值P=(100-A1-A2-A3)*B1+C1+C2，将用户信用等级划分为优秀(P≥100)、良好(80≤P＜100)、中等(50≤P＜80)、差(P<50)。

To implement the construction of the basic credit information database, this dataset collects water consumption data and payment behavior data generated in economic activities, and evaluates users' credit levels via data mining technologies. By monitoring credit assessment data, a risk early warning mechanism is established to timely identify users with potential malicious arrears or water safety hazards, and take targeted measures to ensure water supply safety. 1. Collect historical payment data of users, including total water consumption in the previous year, total expenses in the previous year, total water consumption in the previous month, total annual year-on-year water consumption in the previous year, date of expense generation in the previous month, payment start date, overdue date, and actual payment date. 2. Analyze and extract user credit default characteristic data (number of overdue occurrences A1, maximum overdue days A2, average overdue days A3), on-time payment creditworthy characteristic data (proportion of on-time payments B1), and water use behavior characteristic data (average monthly total water consumption in the previous year, year-on-year difference percentage C1 = (total water consumption in the previous month - total annual water consumption in the previous year / 12) / (total annual water consumption in the previous year / 12), average monthly water consumption difference percentage C2 = (total water consumption in the previous month - average monthly total water consumption in the previous year) / average monthly total water consumption in the previous year). 3. Calculate using the summarized user credit default features, creditworthy features and water use features, with the score formula P = (100 - A1 - A2 - A3) * B1 + C1 + C2, and divide user credit ratings into four levels: excellent (P ≥ 100), good (80 ≤ P < 100), moderate (50 ≤ P < 80), and poor (P < 50).

搜集汇总

数据集介绍

特点

该数据集为宁海县个人用水用户信用评估数据，包含用户的用水量、缴费行为、逾期情况等指标，用于评估用户信用水平。数据集规模为1001条，每月更新，适用于信用信息基础库建设和风险预警机制。