CN-116860820-B - User product preference analysis method and system based on data security

CN116860820BCN 116860820 BCN116860820 BCN 116860820BCN-116860820-B

Abstract

The invention discloses a user product preference analysis method and a system based on data security, wherein the method comprises the following steps: S1, acquiring browsing information and purchasing information of a user, and analyzing to obtain user behavior parameters; the method comprises the steps of S2, cleaning data of obtained user behavior parameters, classifying browsing and purchasing data based on types of products to obtain classified behavior data, S3, calculating subjective preference and objective preference of users to the products according to the classified behavior data, S4, combining the subjective preference and objective preference of the users according to preset credibility, and calculating comprehensive preference of the users to the products, S5, analyzing purchasing preference of the users to the products according to the comprehensive preference, and recommending corresponding products to the users based on the comprehensive preference of the users. The invention classifies the interpretation result of the user product preference value and desensitizes the data, thereby improving the use efficiency of the user product preference analysis method based on data security.

Inventors

TANG HAO
CHEN XIUTING
WANG WEI
LI SHENGQIANG
QIU CHENGQUN

Assignees

盐城师范学院

Dates

Publication Date: 20260508
Application Date: 20230726

Claims (6)

1. A method for analyzing user product preferences based on data security, the method comprising the steps of: S1, acquiring browsing information and purchasing information of a user, and analyzing to obtain user behavior parameters; S2, cleaning the obtained user behavior parameters, and classifying browsing and purchasing data based on the types of products to obtain classified behavior data; S3, calculating subjective preference degree and objective preference degree of the user on the product according to the classified behavior data; the step of calculating the subjective preference degree and the objective preference degree of the user on the product according to the classified behavior data comprises the following steps: S31, classifying behavior data into subjective data and objective data, and performing desensitization treatment on the subjective data and the objective data; S32, calculating objective preference degree of the user according to the desensitized objective data; The step of calculating the objective preference degree of the user according to the desensitized objective data comprises the following steps: s321, grouping the desensitized objective data according to the types of products to obtain objective grouping data; s322, a time sequence prediction model is established according to the grouping result, analysis and prediction are carried out on objective data after grouping according to the time sequence prediction model to obtain a time sequence prediction value, and weighting is carried out on the time sequence prediction value; The method for establishing the time sequence prediction model according to the grouping result comprises the following steps of analyzing and predicting the objective data after grouping according to the time sequence prediction model to obtain a time sequence prediction value, and weighting the time sequence prediction value: s3221, acquiring historical data according to classification behavior data, analyzing the historical data to obtain objective data, and preprocessing the objective data; S3222, constructing and training an improved BP neural network model according to the preprocessed objective data; the construction and training of the improved BP neural network model according to the preprocessed objective data comprises the following steps: s32221, 60% of the preprocessed objective data are used as training sets, and 40% of the preprocessed objective data are used as test sets; s32222, a personalized system recommendation algorithm is selected to construct an improved BP neural network model, and model training is carried out by utilizing the training set, wherein the formula of the improved BP neural network model is as follows: ; in the formula, A time sequence predicted value of objective data; similarity between objective data; Scoring vectors for objective data; Is a mixed weight factor; The number of the scoring values; scoring the packet data a first time; S32223, inputting a testing machine into the improved BP neural network model after training is finished for testing, performing multiple training and testing on the improved BP neural network model by using a cross-validation method, and calculating an average error to estimate the accuracy of the result; s32223, iterating the improved BP neural network model according to the estimation result; S3223, predicting the collected objective data in real time according to the improved BP neural network model; S3224, taking the prediction result of the improved BP neural network model as a time sequence prediction value of objective data; s323, calculating an objective preference value according to the objective grouping data, and giving weight to the objective preference value; S324, calculating objective preference degree of a user on a product based on the weighted time sequence predicted value and the objective preference value; s33, calculating subjective preference degree of the user according to the desensitized subjective data; S4, combining the subjective preference degree and the objective preference degree of the user according to the preset confidence degree, and calculating the comprehensive preference degree of the user on the product; s5, analyzing purchasing preference of the user for the product according to the comprehensive preference degree, and recommending the corresponding product for the user based on the comprehensive preference of the user.
2. The method for analyzing preference of user products based on data security according to claim 1, wherein the classifying behavior data into subjective data and objective data and the desensitizing the subjective data and the objective data comprises the steps of: the subjective data and the objective data after desensitization are classified according to regions and time, and statistical analysis is carried out on the subjective data and the objective data after desensitization, so that statistical indexes of average value, standard deviation, maximum value and minimum value of the subjective data and the objective data after desensitization are obtained.
3. The method for analyzing user product preferences based on data security according to claim 1, wherein the step of calculating subjective preference of the user based on the subjective data after desensitization comprises the steps of: S331, performing data pre-analysis on subjective data; And S332, calculating a subjective preference value according to the subjective data after the pre-analysis, and carrying out weighting calculation on the subjective preference value to obtain the subjective preference.
4. The method for analyzing user product preferences based on data security according to claim 3, wherein the step of combining subjective preference and objective preference of the user according to the preset confidence level and calculating the comprehensive preference of the user to the product comprises the following steps: S41, respectively assigning weights to the subjective preference degree and the objective preference degree; s42, calculating according to the weight of the subjective preference degree and the weight of the objective preference degree to obtain the comprehensive preference degree.
5. The method for analyzing user product preferences based on data security according to claim 4, wherein the steps of analyzing the purchase preferences of the user for the products according to the integrated preferences and recommending the corresponding products to the user based on the integrated preferences of the user comprise the steps of: S51, recommending the same kind of articles according to the articles with higher comprehensive preference of the user products based on the comprehensive preference of the user; s52, searching similar articles according to the comprehensive preference of the user products, improving the pushing frequency of the similar articles, analyzing the real-time prices of the similar articles according to the comprehensive preference of the user products, and pushing real-time price changes to the user.
6. A data security based user product preference analysis system for implementing the steps of the data security based user product preference analysis method of any one of claims 1-5, the system comprising: The data acquisition module is used for acquiring browsing information and purchasing information of a user and analyzing the browsing information and purchasing information to obtain user behavior parameters; The data cleaning module is used for cleaning the data of the obtained user behavior parameters, classifying the browsing and purchasing data based on the types of the products, and obtaining classified behavior data; The behavior data analysis module is used for calculating subjective preference and objective preference of the user on the product according to the classified behavior data; The comprehensive preference degree acquisition module is used for combining the subjective preference degree and the objective preference degree of the user according to the preset confidence degree and calculating the comprehensive preference degree of the user on the product; and the recommendation adjustment module is used for analyzing the purchasing preference of the user for the product according to the comprehensive preference degree and recommending the corresponding product for the user based on the comprehensive preference of the user.

Description

User product preference analysis method and system based on data security Technical Field The invention relates to the technical fields of data analysis and data security, in particular to a user product preference analysis method and system based on data security. Background With the rapid development of the Internet, people are widely applied to the Internet, online shopping is particularly focused by people, the number of servers accessed to the Internet and the number of web pages on the Internet are in an exponentially growing situation, a large amount of information is simultaneously presented in front of the world, the global information amount is increased through digital revolution, the individual has too much information to process, the information utilization rate is lowered due to information explosion, the user product preference is more and more important, the purchasing preference analysis is to mine historical behavior data and predict the future event purchasing behavior of the user according to reliable calculation, the user product preference analysis is a key market research task and aims to know the demands and preferences of different products of consumers, and meanwhile, the data security of the user is very important when the user product preference analysis is carried out. While data security is of increasing interest, data security is in the sense of protecting sensitive information of individuals and organizations from unauthorized access, use, modification, leakage or damage. The data security is important to the protection of finance, intellectual property, customer records, health records and other sensitive information of enterprises, institutions, organizations and individuals, and the effective data security is beneficial to maintaining the reputation and the reputation of the enterprises, organizations and individuals, promoting the service growth and development, reducing the influence of data leakage on the customers and improving the public praise when the products are sold. The existing user preference analysis method is more focused on evaluating the personal consumption level of a user when obtaining information by mining data in a historical browse log of retail products, and relies on real-time browse data when predicting the types, styles and the like of products required by the user, but cannot develop the influence brought by product sales of live broadcast with goods and net red effect bands according to the modern society to perform real-time user product preference analysis, so that the existing user product preference analysis cannot meet the demands of merchants. For the problems in the related art, no effective solution has been proposed at present. Disclosure of Invention Aiming at the problems in the related art, the invention provides a user product preference analysis method and a system based on data security, so as to overcome the technical problems in the prior related art. For this purpose, the invention adopts the following specific technical scheme: a method of user product preference analysis based on data security, the method comprising the steps of: S1, acquiring browsing information and purchasing information of a user, and analyzing to obtain user behavior parameters; S2, cleaning the obtained user behavior parameters, and classifying browsing and purchasing data based on the types of products to obtain classified behavior data; S3, calculating subjective preference degree and objective preference degree of the user on the product according to the classified behavior data; S4, combining the subjective preference degree and the objective preference degree of the user according to the preset confidence degree, and calculating the comprehensive preference degree of the user on the product; s5, analyzing purchasing preference of the user for the product according to the comprehensive preference degree, and recommending the corresponding product for the user based on the comprehensive preference of the user. Further, the calculating the subjective preference and the objective preference of the user to the product according to the classified behavior data includes the following steps: S31, classifying behavior data into subjective data and objective data, and performing desensitization treatment on the subjective data and the objective data; S32, calculating objective preference degree of the user according to the desensitized objective data; s33, calculating subjective preference degree of the user according to the desensitized subjective data. Further, the step of calculating the objective preference of the user according to the objective data after desensitization comprises the following steps: s321, grouping the desensitized objective data according to the types of products to obtain objective grouping data; s322, a time sequence prediction model is established according to the grouping result, analysis and prediction are carried out on objecti