A Classification Method of Customer Group Based on Customer's Wandering Behavior

Abstract

The invention discloses a method for classifying customer groups based on customers' wandering behavior, which includes the following steps: step 1, dividing the shopping mall into regions; step 2, using wifi positioning information to obtain the data of the stay time of customers in each region when they visit the shopping mall each time; Step 3, judge the validity of the data obtained in step 2, and obtain the effective shopping ratio of customers in each area; Step 4, perform principal component processing on the effective shopping ratio, and obtain the customer's feature vector; Step 5, analyze the feature vector Carry out clustering, train and test the decision tree model on the clustering results, and obtain the classification model: step 6, the corresponding test results are different for different categories, process the test results, determine the optimal number of categories, and the corresponding classification model is Customers who visit the mall can be categorized. This method can use the customer's staying behavior information in the mall to better classify customers.

Description

technical field [0001] The invention belongs to the technical field of customer classification, and relates to a method for classifying customer groups in shopping malls, in particular to a method for classifying customer groups based on customer shopping behavior. Background technique [0002] The existing shopping mall customer group classification methods mainly include the following: [0003] 1) Manually distribute questionnaires to ask customers visiting the mall about their time preference and frequency of visits. But the limitation is that a general questionnaire can only obtain information on hundreds of people at most, and the mall needs to arrange a certain budget, which is far from enough to understand the classification of customer groups visiting a large shopping mall. [0004] 2) Classify the customer groups through the member information of the mall, and use the consumption information (including time and consumption amount) of members to group each time. Bu...

AI Technical Summary

Problems solved by technology

But the limitation is that the number of members in general shopping malls will be very small compared to the total number of customers visiting the shopping malls, generally only a few percentage points

Such a sampling rate is far from being able to have a full and accurate understanding of customers visiting the mall

Images