Intelligent rights and interests recommendation method based on customer levels

Abstract

The invention discloses an intelligent rights and interests recommendation method based on customer levels, which relates to the technical field of computers and comprises the steps of customer level evaluation, data collection and analysis, data preprocessing and feature extraction, first intelligent recommendation, recommendation result information collection and analysis and secondary intelligent recommendation. According to the method, recommendation result information is collected and analyzed after first intelligent recommendation, so that opinions and suggestions of users for first right recommendation are widely listened, recommendation is stopped for clients who do not want right recommendation, other right recommendation is performed for users who want other right recommendation, and the user experience is improved. According to the method, richer right recommendation is provided for customers who want to receive right recommendation, and precise recommendation is provided for customers with specified right requirements, so that differentiated recommendation and precise recommendation can be realized, disturbance to some customers is avoided, the favor of the customers to enterprises or platforms is improved as much as possible, and the user experience is improved. And the customer experience effect is better.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method for intelligently recommending rights and interests based on customer levels. Background Art

[0002] With the development of the Internet and the trend of consumption upgrading, the membership system has gradually become a common practice in all walks of life. The membership system can not only increase the loyalty and stickiness of the enterprise, but also provide members with better services and discounts, and improve user experience and satisfaction. Therefore, it is very important for enterprises to design a reasonable and attractive membership rights plan. Membership rights attract more users to join through exclusive benefits and privileges, thereby increasing the company's revenue and market share. In addition, the membership rights plan can also increase user stickiness, improve user loyalty, and make users more willing to choose the company's products and services.

[0003] In order to meet the needs and consumption capabilities of different users, companies usually use computers to recommend membership benefits. By providing more personalized recommendations, they improve user satisfaction and corporate profits. Recommendation systems are widely used in various fields, such as product recommendations on e-commerce platforms, news push on news websites, song recommendations on music platforms, etc. These applications not only improve user experience, but also significantly increase the user stickiness and commercial value of the platform.

[0004] The existing method of recommending corporate benefits is unable to collect user feedback after the recommendation. In actual use, we found that the membership benefits recommended to users by some e-commerce, music and other platforms will cause customers to be disgusted. Due to the lack of the ability to collect customer feedback information, the platform is unable to timely learn about customers' favorability towards the benefits, resulting in the platform often recommending benefits that customers do not need, poor customer experience, and many negative reviews on the platform cause collective rejection by customers, which in turn affects the company's profits and reputation. Summary of the invention

[0005] The purpose of the present invention is to provide a method for intelligently recommending rights and interests based on customer level to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A rights and interests intelligent recommendation method based on customer level, including customer level evaluation, data collection and analysis, data preprocessing and feature extraction, first intelligent recommendation, recommendation result information collection and analysis and second intelligent recommendation, the method steps are as follows;

[0008] S1. Customer level assessment: by collecting the customer's historical consumption data and points accumulation in the past year or half a year, and by screening and classifying these data, the user's consumption field and consumption level can be determined, and then the customer's level can be assessed;

[0009] S2. Data collection and analysis: collecting customers’ historical behavior data and basic information to understand their interests, preferences, backgrounds, and characteristics;

[0010] S3, data preprocessing and feature extraction, preprocess the collected data, and then extract feature vectors that can reflect customer interests and characteristics from the original data;

[0011] S4, first intelligent recommendation, recommending corresponding benefits based on the customer's level and interest preferences;

[0012] S5. Collection and analysis of recommendation result information;

[0013] S6. Secondary intelligent recommendation: Make secondary intelligent recommendations to customers based on the results of recommendation result information collection and analysis. For customers who are successful or satisfied with the recommendation, make the same type of recommendation again. For customers who are unsuccessful with the recommendation and have not provided feedback on whether they are satisfied, recommend other types of products. For customers who are unsuccessful with the recommendation and are not satisfied, stop recommending any products. For customers who are unsuccessful with the recommendation and are not satisfied, extract the customer's reply information and decide whether to recommend and the type of recommendation based on the customer's requirements.

[0014] As a further solution of the present invention: the customer's historical consumption information in step S1 includes the customer's purchase unit price, total purchase amount and number of purchases.

[0015] As a further solution of the present invention: the historical behavior data in step S2 includes browsing records, search records and purchase records.

[0016] As a further solution of the present invention: the basic information of the customer in step S2 includes age, gender and geographical location.

[0017] As a further solution of the present invention: the data preprocessing in step S3 includes data cleaning, deduplication and missing value processing to ensure the quality and integrity of the data.

[0018] As a further solution of the present invention: the feature vector in step S3 includes preference tags, purchase frequency and browsing time.

[0019] As a further solution of the present invention: the rights and interests in step S4 can be distributed to customers in the form of coupons, subsidies, full discounts, flash sales and lotteries, and customers can be stimulated through rights and interests marketing to achieve the purpose of acquiring, activating and retaining customers, thereby increasing business revenue.

[0020] As a further solution of the present invention: the recommendation result information collection in step S5 includes recommendation success rate collection, customer satisfaction collection and customer response information collection, and the customer satisfaction collection and customer response information collection are collected in the form of telephone return visits.

[0021] As a further solution of the present invention: the recommendation method in step S4 and step S6 can be text message, APP notification or phone call, and data anonymization and de-identification, differential privacy, and homomorphic encryption technology are used to protect the user's privacy information.

[0022] As a further solution of the present invention: the first intelligent recommendation in step S4 and the second intelligent recommendation in step S6 are both based on a rights recommendation model constructed using the scrolling neural network technology. The model analyzes the user's consumption habits, predicts the rights that the user may be interested in, and outputs the recommendation results.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] 1. After the first intelligent recommendation, the present invention will collect and analyze the recommendation result information, so as to widely listen to the opinions and suggestions of users on the first benefit recommendation, stop recommending to customers who do not want benefit recommendations, recommend other benefits to users who want other benefit recommendations, give more abundant benefit recommendations to customers who are willing to receive benefit recommendations, and provide precise recommendations to customers with specified benefit requirements. In this way, differentiated recommendations and precise recommendations can be achieved while avoiding disturbing some customers, thereby enhancing customers' favorability towards the enterprise or platform as much as possible and making the customer experience better.

[0025] 2. The present invention protects the user's privacy information by adopting technologies such as data anonymization and de-identification, differential privacy, and homomorphic encryption, and can provide accurate recommendation services while ensuring user privacy. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 The figure is an overall flow chart of a method for intelligent recommendation of benefits based on customer level.

[0027] Figure 2 A method for intelligent recommendation of rights based on customer level Figure 1 The composition structure diagram of secondary intelligent recommendation.

[0028] Figure 3 A method for intelligent recommendation of rights based on customer level Figure 1 Flowchart of secondary intelligent recommendation.

[0029] Figure 4A diagram showing the composition of the data collection structure in a rights intelligent recommendation method based on customer level. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] See also Figure 1-4 In an embodiment of the present invention, a rights and interests intelligent recommendation method based on customer level includes customer level evaluation, data collection and analysis, data preprocessing and feature extraction, first intelligent recommendation, recommendation result information collection and analysis and second intelligent recommendation.

[0032] The method steps are as follows: first, the customer level assessment is carried out by collecting the customer's historical consumption data and points accumulation in the last year or half a year. The customer's historical consumption information includes the customer's purchase unit price, total purchase amount and purchase frequency. By screening and classifying these data, the user's consumption field and consumption level can be determined, and then the customer's level can be assessed;

[0033] Then, data collection and analysis are performed. The customer's historical behavior data, including browsing history, search history, and purchase history, are collected to understand the customer's interests and preferences. The customer's basic information, including age, gender, and geographic location, are collected to understand the customer's background and characteristics.

[0034] Then, data preprocessing and feature extraction are performed. The collected data is preprocessed. Data preprocessing includes data cleaning, deduplication and missing value processing to ensure the quality and integrity of the data. Feature vectors that can reflect customer interests and characteristics are then extracted from the original data. Feature vectors include preference labels, purchase frequency and browsing time.

[0035] Then, we make the first intelligent recommendation, and recommend corresponding benefits according to the customer's level and interest preferences. The benefits can be distributed to customers in the form of coupons, subsidies, discounts, flash sales, and lucky draws. We stimulate customers through benefit marketing to achieve the goals of acquiring, activating, and retaining customers, thereby increasing business revenue.

[0036] Then, the recommendation result information is collected and analyzed. The recommendation result information collection includes the recommendation success rate collection, customer satisfaction collection and customer response information collection. The customer satisfaction collection and customer response information collection are collected through telephone return visits;

[0037] Finally, we conduct a secondary intelligent recommendation. For customers who are successfully recommended or satisfied, we will make the same type of recommendation again. For customers who are unsuccessful in recommendation and have not provided feedback on whether they are satisfied, we will recommend other types of products. For customers who are unsuccessful in recommendation and are not satisfied, we will stop recommending any products. After extracting the customer’s reply information, we will decide whether to recommend and the type of recommendation based on the customer’s requirements.

[0038] The recommendation method in the first smart recommendation and the second smart recommendation can be SMS, APP notification or phone call. The recommendation system is one of the core businesses of modern Internet companies. It provides users with personalized product, service and content recommendations by analyzing user behavior, interests and preferences. The first smart recommendation and the second smart recommendation are both based on the rights recommendation model built using the scrolling neural network technology. The model analyzes the user's consumption habits, predicts the rights that the user may be interested in, and outputs the recommendation results.

[0039] As the scale of data increases, the accuracy and effectiveness of the recommendation system have also been significantly improved. However, this has also brought about a series of ethical and privacy issues. Based on this, the present invention adopts data anonymization and de-identification, differential privacy, and homomorphic encryption technology to protect the user's privacy information, ensuring that accurate recommendation services are provided while protecting user privacy.

[0040] The present invention will collect and analyze the recommendation result information after the first intelligent recommendation, and provide differentiated and precise recommendations to customers based on the analysis, while reducing disturbances to some customers who do not need recommendations and reducing the impact on customers' lives, thereby improving these customers' views on the company or platform, and enhancing customers' favorability towards the company or platform as much as possible, resulting in a better customer experience.

[0041] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A rights and interests intelligent recommendation method based on customer level, including customer level evaluation, data collection and analysis, data preprocessing and feature extraction, first intelligent recommendation, recommendation result information collection and analysis and second intelligent recommendation, characterized in that: The method steps are as follows: S1. Customer level assessment: by collecting the customer's historical consumption data and points accumulation in the past year or half a year, the customer's level is assessed; S2. Data collection and analysis: collecting customers’ historical behavior data and basic information to understand their interests, preferences, backgrounds, and characteristics; S3, data preprocessing and feature extraction, preprocess the collected data, and then extract feature vectors that can reflect customer interests and characteristics from the original data; S4, first intelligent recommendation, recommending corresponding benefits based on the customer's level and interest preferences; S5. Collection and analysis of recommendation result information; S6. Secondary intelligent recommendation: Make secondary intelligent recommendations to customers based on the results of recommendation result information collection and analysis. For customers who are successful or satisfied with the recommendation, make the same type of recommendation again. For customers who are unsuccessful with the recommendation and have not provided feedback on whether they are satisfied, recommend other types of products. For customers who are unsuccessful with the recommendation and are not satisfied, stop recommending any products. For customers who are unsuccessful with the recommendation and are not satisfied, extract the customer's reply information and decide whether to recommend and the type of recommendation based on the customer's requirements.

2. According to claim 1, it is characterized in that: The customer's historical consumption information in step S1 includes the customer's purchase unit price, total purchase amount and purchase times.

3. According to claim 1, it is characterized in that: The historical behavior data in step S2 includes browsing history, search history and purchase history.

4. The method according to claim 1, characterized in that: The basic information of the customer in step S2 includes age, gender and geographical location.

5. The method according to claim 1, characterized in that: The data preprocessing in step S3 includes data cleaning, deduplication and missing value processing.

6. The method according to claim 1, characterized in that: The feature vector in step S3 includes preference tags, purchase frequency and browsing time.

7. The method according to claim 1, characterized in that: The benefits in step S4 can be distributed to customers in the form of coupons, subsidies, full discounts, flash sales, and lucky draws. Customers can be stimulated through benefit marketing to achieve the purpose of acquiring, activating, and retaining customers, thereby increasing business revenue.

8. The method according to claim 1, characterized in that: The recommendation result information collection in step S5 includes recommendation success rate collection, customer satisfaction collection and customer response information collection. The customer satisfaction collection and customer response information collection are collected in the form of telephone return visits.

9. The method according to claim 1, characterized in that: The recommended method in step S4 and step S6 can be text message, APP notification or phone call, and data anonymization and de-identification, differential privacy, and homomorphic encryption technology are used to protect the user's privacy information.

10. The method according to claim 1, characterized in that: The first intelligent recommendation in step S4 and the second intelligent recommendation in step S6 are both made by constructing a rights recommendation model using the curling neural network technology. The model analyzes the user's consumption habits, predicts the rights that the user may be interested in, and outputs the recommendation results.

Images