The analysis process here is similar to the analysis of business bulk sms service requirements for embedded points. For details, you can view the analysis of business requirements for embedded points, which will not be expanded here. Step 2: Build a label system From the point of view of the data product construction label system, labels can be divided into bulk sms service three categories according to the statistical methods of labels: statistical labels, rule labels, and prediction labels; 1. Statistical labels User portraits are the most basic and common labels. Statistical label values count user-related values and labels that objectively describe user status.
This type of data can usually be obtained from bulk sms service user registration data, user access, and consumption statistics. For example, for a user, its gender, age, city, constellation, active days in the past 7 days, active times in the past 7 days, cumulative purchase amount, cumulative purchase number, monthly average consumption amount and other fields, 2. Rule class label Based on user behavior and determined rules, in the process of actual development bulk sms service of portraits, according to business needs, operators and data personnel jointly negotiate and formulate, including activity tags, RFM tags, etc. For example, the definition of "active transactions" on the platform is "the number of transactions in the past 90 days> 3".
The following is the introduction of the straw hat boy, the bulk sms service commonly used user activity label, the division method of RFM label. (1) User activity label In actual business scenarios, it will involve labeling users with high activity, medium activity, low activity, and churn according to their activity. How is the time range of high, high and low divided in this process? Here, forehead slap is not feasible, and the construction of labels should be based on definitions bulk sms service and methods. First, divide the user's churn cycle. There are usually two methods: One is the inflection point theory: an increase in the value on the X-axis will bring about a substantial gain on the Y-axis, until after a certain point, when X increases