User grouping method and device based on RFM model and readable medium

Abstract

The invention discloses a user grouping method and device based on an RFM model and a readable medium, and the method comprises the steps: obtaining order data of a user, and carrying out the data cleaning of the order data of the user; according to the order data after data cleaning, calculating three indexes, namely a time interval R of last order placing of the user, the order placing frequency F in a specified time period and the order placing total amount M in the specified time period; setting index sub-boxes based on the numerical value intervals of the time interval R, the order placing frequency F and the order placing total amount M, and normalizing the time interval R, the order placing frequency F and the order placing total amount M through the index sub-boxes to obtain corresponding index values RS, FS and MS; determining weight coefficients corresponding to the indexes by adopting an entropy weight method, and calculating a final score of each index according to the index values RS, FS and MS and the weight coefficients; and inputting the final score of each index into a K-means clustering algorithm to obtain an optimal grouping result of the users. Data support is provided for enterprise operation, decision making and project stage summarization.

Description

technical field [0001] The invention relates to the field of data classification, in particular to a user grouping method, device and readable medium based on an RFM model. Background technique [0002] With the transformation of the marketing concept of modern enterprises, from the previous "product-centric" to the current "customer-centric", users have become an increasingly important resource for enterprises. Enterprises formulate different service plans for users at different stages. From the above, it can be seen that scientifically dividing user groups, formulating corresponding user service strategies, and providing differentiated user service strategies to achieve personalized services are important ways for enterprises to maximize the benefits of limited resources and help In order for enterprises to occupy an advantageous position in the fierce market competition. [0003] At present, the RFM model is widely used in the field of user segmentation, but the user se...

AI Technical Summary

Problems solved by technology

[0003] At present, the RFM model is widely used in the field of user segmentation, but the user segmentation of the traditional RFM model is mainly based on the average value of the three indicators R, F, and M to directly divide users into 8 categories, which often leads to ambiguity in user classification.

K-means clustering analysis algorithm, as a classic iterative solution partition clustering algorithm, is often used in combination with RFM model to improve user classification accuracy. However, the traditional K-means clustering analysis algorithm has certain limitations. Manual experience setting, and the estimation of k value is very difficult; the initial clustering center is randomly selected, which may easily lead to inaccurate clustering results; the existence of isolated points may easily increase the number of clustering iterations, and the clustering will fall into local optimum; When there are many types of data attribute values ​​and the importance is inconsistent, the Euclidean distance used treats the importance weight of each attribute with the same weight, resulting in deviations in clustering accuracy

[0004] Secondly, in the RFM model, a good index weight is the key to the performance of the RFM model, but in most RFM model applications, the weights of each index used to identify user value are the same, which has a negative impact on the accuracy of user value evaluation

Images