User portrait generation method and system for medical beauty users

Abstract

The present application relates to the technical field of electric digital data processing, in particular to a user portrait generation method and system for medical beauty users, which divides medical beauty users into groups based on the differences in the consumption data of the medical beauty users, obtaining a plurality of first medical beauty user groups; analyzes the multi-platform medical beauty trend based on the medical beauty trend data of the medical beauty platform and in combination with the consumption data of different medical beauty users in each first medical beauty user group, determines the medical beauty characteristic trend penetration of each first medical beauty user group, and further determines the influence degree of the medical beauty behavior of each medical beauty user on the trend; divides the medical beauty users into groups again based on the differences between the medical beauty behavior influence degrees of different medical beauty users, obtaining a plurality of second medical beauty user groups; and generates the user portrait of each medical beauty user based on the medical beauty characteristic labels corresponding to the two medical beauty user groups in which each medical beauty user is located. The present application improves the accuracy of generating the user medical beauty portrait.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric digital data processing, in particular to a user portrait generation method and system for medical beauty users. BACKGROUND

[0002] With the upgrading of consumption and the growing demand for beauty, the medical beauty industry has shown explosive growth in recent years. Under the trend of increasingly diversified user demand in the medical beauty industry, traditional static labels have been difficult to meet the demand for precise service. With the deep application of big data and artificial intelligence technology, the industry is accelerating the construction of a dynamic user portrait system to achieve in-depth insight into customer demand through real-time data mining.

[0003] Traditional methods such as the RFM model mainly rely on historical consumption data or one-time face-to-face diagnosis to generate static labels, which is difficult to capture user dynamic behavior in real time, and thus cannot effectively understand the changing trends behind user demand and predict future interest directions. In particular, the traditional RFM model only focuses on measuring the past consumption ability of customers, reflecting the historical value of customers, ignoring the dynamic changes of user behavior influenced by popular medical trends, and lacking sharp perception of user potential demand and trends, resulting in inaccurate prediction of new projects or emerging needs that users may be interested in, and thus the generated user portrait is not accurate enough. SUMMARY

[0004] To solve the above technical problems, the purpose of the present application is to provide a user portrait generation method and system for medical beauty users, and the technical solution adopted is as follows:

[0005] In a first aspect, the present application provides a user portrait generation method for medical beauty users, comprising the following steps:

[0006] Obtain consumption data of different medical beauty users and medical beauty popular data of different medical beauty platforms;

[0007] Based on the differences between the consumption data of different medical beauty users, the first group division is performed on all medical beauty users to obtain a plurality of first medical beauty user groups;

[0008] Based on the medical beauty popular data of different medical beauty platforms and combined with the consumption data of different medical beauty users in each first medical beauty user group, the multi-platform medical beauty popular trend analysis is performed to determine the medical feature popular trend penetration of each first medical beauty user group;

[0009] Based on the consumption data of each medical beauty user and combined with the medical beauty popular data of the different medical beauty platforms, the popular medical beauty dynamic characteristic analysis is performed on each medical beauty user, and combined with the medical feature popular trend penetration of the first medical beauty user group where each medical beauty user is located, the medical behavior influence degree of each medical beauty user under the influence of popular trends is determined.

[0010] Based on the difference between the medical beauty behavior influence degrees of different medical beauty users, all the medical beauty users are divided into a second medical beauty user group, and a plurality of second medical beauty user groups are obtained.

[0011] Based on the medical beauty feature labels corresponding to the first medical beauty user group and the second medical beauty user group to which each medical beauty user belongs, a user portrait of each medical beauty user is generated.

[0012] In combination with the first aspect, in some possible implementation manners, determining the medical beauty feature trend penetration degree of each first medical beauty user group comprises:

[0013] Based on the occurrence frequency of the medical beauty keywords in the recent medical beauty popular data of each medical beauty platform and the frequency of co-occurrence of each medical beauty keyword and other medical beauty keywords in the same information source in each medical beauty platform, determining the linkage keywords of each medical beauty keyword on each medical beauty platform and the comprehensive high-frequency linkage degree between each medical beauty keyword and its linkage keywords;

[0014] Based on the difference between the linkage keywords of each medical beauty keyword on different medical beauty platforms and the comprehensive high-frequency linkage degree between each medical beauty keyword and its linkage keywords, determining the multi-platform medical beauty feature trend degree of each medical beauty keyword;

[0015] Based on the multi-platform medical beauty feature trend degree of each medical beauty keyword and in combination with the correlation between all the medical beauty keywords involved in the recent medical beauty projects of all the medical beauty users in each first medical beauty user group and the linkage keywords of each medical beauty keyword, determining the medical beauty feature trend penetration degree of the recent medical beauty projects of each first medical beauty user group and each medical beauty keyword;

[0016] Based on the medical beauty feature trend penetration degree of the recent medical beauty projects of each first medical beauty user group and all the medical beauty keywords, determining the medical beauty feature trend penetration degree of each first medical beauty user group.

[0017] In combination with the first aspect, in some possible implementation manners, determining the linkage keywords of each medical beauty keyword on each medical beauty platform and the comprehensive high-frequency linkage degree between each medical beauty keyword and its linkage keywords comprises:

[0018] Based on the occurrence frequency of the medical beauty keywords in the recent medical beauty popular data of each medical beauty platform, determining the high-frequency degree of each medical beauty keyword;

[0019] Based on the frequency of co-occurrence of each medical beauty keyword and other medical beauty keywords in the same information source in each medical beauty platform, determining the linkage occurrence coefficient of each medical beauty keyword and other medical beauty keywords.

[0020] Based on the frequency and correlation coefficient of each medical aesthetics keyword with other medical aesthetics keywords, the high frequency correlation of each medical aesthetics keyword with other medical aesthetics keywords is determined.

[0021] Based on the aforementioned high-frequency correlation, the associated keywords for each medical aesthetics keyword on each medical aesthetics platform are determined from all other medical aesthetics keywords.

[0022] Based on the high-frequency correlation between each medical aesthetics keyword and its associated keywords, the comprehensive high-frequency correlation between each medical aesthetics keyword and its associated keywords on each medical aesthetics platform is determined.

[0023] In conjunction with the first aspect mentioned above, among some possible implementation methods, determining the multi-platform popularity trend of each medical aesthetics keyword includes:

[0024] Determine the intersection of each medical aesthetics keyword with its associated keywords on different medical aesthetics platforms to obtain the multi-platform associated medical aesthetics keywords for each medical aesthetics keyword;

[0025] Based on the correlation between each linked keyword of each medical aesthetics keyword on each medical aesthetics platform and the linked medical aesthetics keywords of each multi-platform, the contribution of each linked keyword of each medical aesthetics keyword on each medical aesthetics platform to the popularity of medical aesthetics features is determined.

[0026] Based on the comprehensive high-frequency correlation and the contribution of the popularity of each medical aesthetics keyword on various medical aesthetics platforms, the popularity trend of each medical aesthetics keyword across multiple platforms is determined.

[0027] In conjunction with the first aspect mentioned above, among some possible implementation methods, the analysis of the dynamic characteristics of popular cosmetic procedures for each user includes:

[0028] Based on the consumption data of each medical aesthetics user, the distribution of medical aesthetics projects is analyzed, the pursuit of popular medical aesthetics projects by medical aesthetics users is analyzed, and the degree of pursuit of popular medical aesthetics techniques by each medical aesthetics user is determined.

[0029] Based on each medical aesthetics user's consumption data, the types of medical aesthetics projects are identified, and combined with recent medical aesthetics trend data from different medical aesthetics platforms, the suitability of popular medical aesthetics projects for medical aesthetics users is analyzed to determine the individual suitability of each medical aesthetics user for popular medical aesthetics projects.

[0030] Based on the popularity of certain cosmetic medical technologies and the individual's suitability for those technologies, the dynamic characteristics of each cosmetic medical user are obtained.

[0031] In conjunction with the first aspect mentioned above, among some possible implementation methods, the degree of pursuit of popular cosmetic techniques by each cosmetic user can be determined, including:

[0032] determine an actual degree of each medical aesthetic project of each medical aesthetic user based on the number of times of each medical aesthetic project in the consumption data of each medical aesthetic user;

[0033] determine a real involvement degree of medical aesthetic projects of each medical aesthetic user based on the distribution level of the actual degree of each medical aesthetic project of each medical aesthetic user;

[0034] determine a popular medical aesthetic technology pursuit degree of each medical aesthetic user based on the number of types of medical aesthetic projects in the consumption data of each medical aesthetic user, and in combination with the real involvement degree of medical aesthetic projects of each medical aesthetic user.

[0035] In combination with the first aspect, in some possible implementation manners, determining the individual popular medical aesthetic adaptation degree of each medical aesthetic user comprises:

[0036] constructing an individual medical aesthetic keyword set of each medical aesthetic user based on medical aesthetic keywords in all medical aesthetic projects in the consumption data of each medical aesthetic user;

[0037] determining the individual popular medical aesthetic adaptation degree of each medical aesthetic user based on the multi-platform medical aesthetic feature popular trend degree of all medical aesthetic keywords in the individual medical aesthetic keyword set.

[0038] In combination with the first aspect, in some possible implementation manners, determining the medical aesthetic behavior influence degree of each medical aesthetic user by the popular trend comprises:

[0039] determining the weight of the popular medical aesthetic technology pursuit degree and the individual popular medical aesthetic adaptation degree of each medical aesthetic user based on the medical aesthetic feature popular trend penetration degree of the first medical aesthetic user group in which each medical aesthetic user is located;

[0040] performing weighted addition on the popular medical aesthetic technology pursuit degree and the individual popular medical aesthetic adaptation degree of each medical aesthetic user by using the weight, to obtain the medical aesthetic behavior influence degree of each medical aesthetic user by the popular trend.

[0041] In combination with the first aspect, in some possible implementation manners, performing first group division on different medical aesthetic users to obtain a plurality of first medical aesthetic user groups comprises:

[0042] determining a first difference value based on the difference between the consumption amounts of medical aesthetic projects in the consumption data of any two medical aesthetic users;

[0043] determining a second difference value based on the difference between the frequency values of each medical aesthetic project type in the consumption data of any medical aesthetic user;

[0044] determine a third difference value based on a difference between medical aesthetic care parts of a medical aesthetic project in consumption data of any one medical aesthetic user;

[0045] determine a multi-dimensional feature distance between any two medical aesthetic users based on the first difference value, the second difference value and the third difference value;

[0046] cluster all medical aesthetic users by using a clustering algorithm based on the multi-dimensional feature distance, and take a plurality of clustering clusters obtained by clustering as a plurality of first medical aesthetic user groups.

[0047] In a second aspect, the present application further provides a user portrait generation system for medical aesthetic users, comprising a memory and a processor. The memory is used to store executable computer program codes, and the processor is used to call and run the executable computer program codes from the memory, so that the system executes the method in the first aspect or any one of the possible implementation manners of the first aspect.

[0048] In a third aspect, the present application further provides a user portrait generation device for medical aesthetic users, and the device comprises:

[0049] a data acquisition module, used to acquire consumption data of different medical aesthetic users and medical aesthetic popular data of different medical aesthetic platforms;

[0050] a first classification module, used to perform first group division on all medical aesthetic users based on a difference between the consumption data of different medical aesthetic users, and obtain a plurality of first medical aesthetic user groups;

[0051] a penetration degree analysis module, used to perform multi-platform medical aesthetic popular trend analysis based on the medical aesthetic popular data of different medical aesthetic platforms and in combination with the consumption data of different medical aesthetic users in each first medical aesthetic user group, and determine a medical aesthetic feature popular trend penetration degree of each first medical aesthetic user group;

[0052] an influence degree analysis module, used to perform popular medical aesthetic self-dynamic characteristic analysis on each medical aesthetic user based on the consumption data of each medical aesthetic user and in combination with the medical aesthetic popular data of the different medical aesthetic platforms, and in combination with the medical aesthetic feature popular trend penetration degree of the first medical aesthetic user group where each medical aesthetic user is located, determine a medical aesthetic behavior influence degree of each medical aesthetic user on a popular trend;

[0053] a second classification module, used to perform second group division on all medical aesthetic users based on a difference between medical aesthetic behavior influence degrees of different medical aesthetic users, and obtain a plurality of second medical aesthetic user groups;

[0054] a portrait generation module, used to generate a user portrait of each medical aesthetic user based on medical aesthetic feature labels corresponding to the first medical aesthetic user group and the second medical aesthetic user group where each medical aesthetic user is located.

[0055] In a fourth aspect, the present application also provides a computer program product, which comprises computer program codes, when the computer program codes are run on a computer, the computer is caused to execute the medical and beauty user portrait generation method in the first aspect or any possible implementation manner of the first aspect.

[0056] In a fifth aspect, the present application also provides a computer readable storage medium, which stores computer program codes, when the computer program codes are run on a computer, the computer is caused to execute the medical and beauty user portrait generation method in the first aspect or any possible implementation manner of the first aspect.

[0057] The present application has the following beneficial effects: the present application integrates the internal consumption data of medical and beauty users and the medical and beauty popular data of different medical and beauty platforms, and on the basis of the group division of medical and beauty users according to the differences between the consumption data of different medical and beauty users to obtain a plurality of first medical and beauty user groups, the medical and beauty popular trend analysis of multiple platforms is carried out based on the medical and beauty popular data of different medical and beauty platforms and combined with the consumption data of different medical and beauty users in each first medical and beauty user group, the medical and beauty characteristic popular trend penetration of the medical and beauty user group under the influence of the medical and beauty characteristic dynamic popular trend is determined, and then the popular medical and beauty dynamic characteristics of each medical and beauty user are analyzed, and the analysis result is combined with the medical and beauty characteristic popular trend penetration of the first medical and beauty user group where each medical and beauty user is located, and the medical and beauty behavior influence degree of each medical and beauty user to the popular trend is determined, so that the medical and beauty user is divided into groups again based on the differences between the medical and beauty behavior influence degrees of different medical and beauty users, and a plurality of second medical and beauty user groups are obtained, so as to realize the capture of the medical and beauty behavior change of medical and beauty users, and finally, the medical and beauty characteristic labels corresponding to the first medical and beauty user group and the second medical and beauty user group where each medical and beauty user is located are used to generate the user portrait of each medical and beauty user. Through analyzing the medical and beauty behavior influence degree of each medical and beauty user to the popular trend, more perfect medical and beauty characteristic labels of users can be obtained, the trend update and high-precision prediction of the medical and beauty user portrait are realized, the ability to find out the potential demand and emerging projects of users is enhanced, and the generation accuracy of the user portrait is improved. BRIEF DESCRIPTION OF DRAWINGS

[0058] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the drawings needed in the following embodiment or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0059] Figure 1 A step flow chart of a user portrait generation method for medical and beauty users according to an embodiment of the present application;

[0060] Figure 2 A structural schematic diagram of a user portrait generation system for medical and beauty users according to an embodiment of the present application;

[0061] Figure 3 A structural schematic diagram of a user portrait generation device for medical and beauty users according to an embodiment of the present application. DETAILED DESCRIPTION

[0062] To clearly illustrate the technical features of the present application, the following will describe the present application in detail with specific embodiments and in conjunction with the drawings.

[0063] Embodiments of the present application will be described below in greater detail with reference to the accompanying drawings. While certain embodiments of the present application are shown in the drawings, it is understood that the present application can be embodied in various forms and should not be construed as being limited to the embodiments set forth herein, but rather, these embodiments are provided so that the present application can be more thoroughly and completely understood. It should be understood that the drawings and embodiments of the present application are only for illustrative purposes and should not be construed as limiting the scope of the present application.

[0064] It should be understood that each of the steps described in the method embodiments of the present application can be performed in different orders and / or in parallel. In addition, the method embodiments can include additional steps and / or omit the execution of the steps shown. The scope of the present application is not limited in this respect.

[0065] The term “comprising” and variations thereof as used herein are open-ended, that is, “including but not limited to”. The term “based on” is “based, at least in part, on”. The term “one embodiment” means “at least one embodiment”; the term “another embodiment” means “at least one additional embodiment”; the term “some embodiments” means “at least some embodiments”. Related definitions will be given in the description below.

[0066] It should be noted that the terms “first”, “second”, and the like used in the present application are only used to distinguish different devices, modules or units, and are not intended to limit the order or interdependence of the functions performed by these devices, modules or units.

[0067] In the embodiments of the present application, although the operations or steps are described in a specific order in the accompanying drawings, it should not be understood as requiring the operations or steps to be performed in the specific order or serial order shown, or requiring all of the operations or steps to be performed to obtain the desired results. In the embodiments of the present application, these operations or steps can be performed in series; they can also be performed in parallel; and a part of them can be performed.

[0068] Meanwhile, it can be understood that the data involved in the technical solutions of the present application (including but not limited to the data itself, the acquisition or use of the data) should comply with the requirements of the corresponding laws and regulations and relevant provisions. Unless otherwise defined, all technical and scientific terms used in the present application have the same meaning as understood by those skilled in the art to which the present application belongs, and all parameters or indicators in the formulas involved in the present application are normalized values that eliminate the influence of dimensions.

[0069] The medical and beauty user portrait generation method and system provided by the embodiment of the present application will be described in detail below with reference to the accompanying drawings.

[0070] Figure 1 The basic flow diagram of the medical and beauty user portrait generation method provided by the embodiment of the present application is shown in FIG. 1, which specifically includes the following steps: Figure 1

[0071] Step S100: Obtain the consumption data of different medical and beauty users and the medical and beauty popular data of different medical and beauty platforms.

[0072] The long-term consumption records and consumption habits of medical and beauty users are helpful to explore potential consumption trends and preferences, so the consumption data of different medical and beauty users can be obtained by means of an order system and a database. The consumption data, also known as internal data of medical and beauty users, refers to the historical consumption transaction data of different medical and beauty users. In the embodiment of the present application, the consumption data of different medical and beauty users includes: project categories (such as injection, skin, surgery, etc.), project tags (such as hydration, whitening, anti-aging, tightening, etc.), consumption amount, consumption time, discount information, etc.

[0073] At the same time, the medical and beauty popular data of different medical and beauty platforms is obtained. The medical and beauty popular data, also known as external trend data, refers to the public data of each medical and beauty platform. In the embodiment of the present application, the medical and beauty popular data of different medical and beauty platforms includes: (1) Project / keyword sound volume: the frequency and growth rate of a specific medical and beauty project (such as “photon skin rejuvenation”, “fotona 4D”, “golden microneedle”) being mentioned, discussed, liked, and collected; (2) Hot notes / posts: title, text keywords, tags (Hashtag), interaction amount (like, collect, comment number), etc.

[0074] At this point, the consumption data of different medical and beauty users and the medical and beauty popular data of different medical and beauty platforms are obtained to facilitate subsequent operations.

[0075] Step S200: Based on the differences between the consumption data of different medical and beauty users, the first group division is performed on all medical and beauty users to obtain a plurality of first medical and beauty user groups.​

[0076] The traditional RFM model usually starts from a single value perspective to divide the medical beauty users into groups, mainly focusing on dimensions such as consumption ability, consumption frequency and consumption amount. However, the user demand of the medical beauty industry is complex and diverse, and a single value perspective cannot fully reflect its characteristics. Therefore, more dimensions should be introduced when dividing the groups of medical beauty users. In addition to the traditional consumption ability and frequency, the user's consumption preferences should also be considered, such as the type of selected projects (anti-aging, hydration, light medical beauty, etc.), nursing site, consumption time period, and specific needs for medical beauty services. Through these dimensions, more business meaningful and practically valuable initial labels can be given to users, such as "high-end anti-aging pursuers", "high-frequency basic nursing customer groups", "potential eye plastic surgery interest groups", etc. These labels not only accurately reflect the user's demand and behavior characteristics, but also provide a more powerful benchmark image for subsequent user analysis and precision marketing.

[0077] Therefore, the above step S200 first divides all medical beauty users into groups based on the differences between the consumption data of different medical beauty users, such as the differences between multiple dimensions such as consumption ability, frequency of each medical beauty project type and medical beauty site in the consumption data, obtains a plurality of medical beauty user groups, and refers to the first medical beauty user group.

[0078] Further, the above step S200 first divides all medical beauty users into groups, obtaining a plurality of first medical beauty user groups, including: determining a first difference value based on the difference between the consumption amount of medical beauty projects in the consumption data of any two medical beauty users; determining a second difference value based on the difference between the frequency value of each medical beauty project type in the consumption data of any medical beauty user; determining a third difference value based on the difference between the medical beauty nursing site of medical beauty projects in the consumption data of any medical beauty user; determining the multi-dimensional feature distance between any two medical beauty users based on the first difference value, the second difference value and the third difference value; and clustering all medical beauty users using a clustering algorithm based on the multi-dimensional feature distance, and taking the plurality of clustering clusters obtained by clustering as the plurality of first medical beauty user groups.

[0079] In the embodiment of the present application, first, the consumption amount of all times in the consumption data of each medical beauty user is averaged, and the average value is normalized by using a normalization function (such as a maximum-minimum normalization function), so as to obtain the consumption ability degree of each medical beauty user.

[0080] Secondly, based on the medical beauty project type in the consumption data of each medical beauty user, a fixed order is arranged, that is, the types of anti-aging, moisturizing, light medical beauty, etc. are sequentially arranged, so as to form a fixed medical beauty project type collection. Then, the number of each medical beauty project type performed by each medical beauty user is normalized by using a normalization function (such as a maximum-minimum normalization function), so as to obtain the frequency value of each medical beauty project type of each medical beauty user; and the frequency value of each medical beauty project type of the user is mapped to the fixed medical beauty project type collection of each medical beauty user one by one, so that each medical beauty project type has its corresponding frequency value.

[0081] Next, based on the medical beauty care parts in the consumption data of each medical beauty user, a medical beauty care part set of each medical beauty user is constructed.

[0082] Finally, based on the above-mentioned determined consumption ability degree, frequency value corresponding to the medical beauty project type, and three dimensions corresponding to the medical beauty care part of each medical beauty user, clustering operation is performed on all medical beauty users. Before the clustering operation, the measurement method of the multidimensional feature distance between any two medical beauty users is defined, that is:

[0083] For the consumption ability degree dimension of the medical beauty user, the consumption ability degrees of any two medical beauty users are subtracted and the absolute value is taken, so as to determine the consumption ability difference between any two medical beauty users, also called the first difference value;

[0084] For the medical beauty project type and the frequency value of the medical beauty user, the corresponding frequency values of each medical beauty project type of any two medical beauty users are subtracted and the absolute value is taken, then the whole is accumulated, and normalized by using a normalization function (such as a maximum-minimum normalization function), so as to determine the medical beauty project type difference between any two medical beauty users, also called the second difference value;

[0085] For the medical beauty care part set of the medical beauty user, the correlation coefficient of the medical beauty care part set of any two medical beauty users is calculated by using the jaccard correlation coefficient, and the difference between the value 1 and the correlation coefficient is calculated, so as to determine the medical beauty care part difference between any two medical beauty users, also called the third difference value;

[0086] In this way, the multidimensional feature distance between any two medical beauty users is calculated by using the calculation method of the Euclidean distance for the above-mentioned three dimension difference values.

[0087] Thus, based on the multi-dimensional feature distance between any two medical beauty users, all medical beauty users are clustered by a DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering method to determine a plurality of medical beauty user groups, which are referred to as first medical beauty user groups. Each first medical beauty user group can represent the performance of medical beauty user features in three different dimensions, and each dimension can represent the specific feature performance of the user group in that dimension.

[0088] A plurality of historical medical beauty user groups and their consumption data are obtained in advance, and a medical beauty feature label is manually formulated for each historical medical beauty user group based on the above-mentioned consumption ability degree, frequency value corresponding to the medical beauty project type, and information corresponding to the three dimensions of the medical beauty care part in the consumption data of each user in each historical medical beauty user group, such as high net worth-high frequency maintenance-face refinement type, rational practical-node ritual-skin management type, and tasting experience-low frequency attempt-face care type.

[0089] Based on the above-mentioned consumption ability degree, frequency value corresponding to the medical beauty project type, and information corresponding to the three dimensions of the medical beauty care part in the consumption data, each first medical beauty user group is matched with each historical medical beauty user group to determine the most matched historical medical beauty user group for each first medical beauty user group. Since the specific implementation process of data matching is well known to those skilled in the art, it will not be described here. The feature label of the most matched historical medical beauty user group of each first medical beauty user group is taken as its own medical beauty feature label, and thus the first feature label of each first medical beauty user group can be obtained for subsequent use.

[0090] Step S300: Based on the medical beauty popular data of different medical beauty platforms and combined with the consumption data of different medical beauty users in each first medical beauty user group, a multi-platform medical beauty popular trend analysis is performed to determine the medical beauty feature popular trend penetration of each first medical beauty user group.

[0091] The step S200 analyzes the consumption data of different medical beauty users and divides all medical beauty users into groups, and assigns each medical beauty user group with specific feature labels, which are static features generated based on the current consumption data of the medical beauty users. However, due to the characteristics of the medical beauty industry, its development has a certain trend, and with the change of people's aesthetic concept and the update of personal medical beauty needs, the user demand will also change. This change not only reflects the diversification of demand, but also may trigger the emergence of new submarkets and potential demand. Therefore, in order to more accurately capture user demand and grasp the future development trend of the medical beauty market, these dynamic factors need to be considered, rather than relying solely on static label analysis.

[0092] Therefore, the step S300 analyzes the medical beauty trend of different medical beauty platforms, and combines the consumption data of different medical beauty users in each first medical beauty user group to determine the medical beauty feature trend penetration of each first medical beauty user group.

[0093] Further, the step S300 of determining the medical beauty feature trend penetration of each first medical beauty user group comprises:

[0094] Step S301: Based on the appearance frequency of medical beauty keywords in the recent medical beauty trend data of each medical beauty platform, and the frequency of each medical beauty keyword and other medical beauty keywords appearing together in the same information source in each medical beauty platform, determine the linkage keywords of each medical beauty keyword on each medical beauty platform, and the comprehensive high-frequency linkage degree between each medical beauty keyword and its linkage keywords.

[0095] In the medical beauty industry, the extension of the trend does not come from a single medical beauty project, but from the coordinated development and evolution of multiple related projects or technologies. The interaction of different technology modules promotes the overall trend of the industry, and also leads to the overlap and interweaving of high-frequency keywords in the content. In this process, medical beauty technologies continue to innovate and integrate, promoting the diversification and in-depth development of the industry. Therefore, when different high-frequency medical beauty keywords frequently appear in the same content source, it means that they have higher relevance and linkage under the current trend, reflecting the close relationship between technology and demand.

[0096] Therefore, the related feature nouns in the medical beauty project type, medical beauty care part, medical beauty technology, etc. of the medical beauty trend data of the medical beauty platform are collectively used as keywords of medical beauty related words, thereby constructing a medical beauty keyword set. When a certain medical beauty keyword in the medical beauty keyword set and other keywords appear frequently in the same information source of each medical beauty platform, it means that the other keyword is a linkage keyword of the certain medical beauty keyword, and the comprehensive high-frequency linkage degree between the certain medical beauty keyword and the linkage keyword is high.

[0097] Further, the determination of the linkage keywords of each medical aesthetic keyword on each medical aesthetic platform and the comprehensive high-frequency linkage degree between each medical aesthetic keyword and its respective linkage keywords in step S301 comprises: determining the high frequency of each medical aesthetic keyword based on the appearance frequency of the medical aesthetic keyword in the recent medical aesthetic popular data of each medical aesthetic platform; determining the linkage appearance coefficient of each medical aesthetic keyword and other medical aesthetic keywords based on the frequency of the co-appearance of each medical aesthetic keyword and other medical aesthetic keywords in the same information source of each medical aesthetic platform; determining the high-frequency linkage degree of each medical aesthetic keyword and other medical aesthetic keywords based on the high frequency and the linkage appearance coefficient of each medical aesthetic keyword and other medical aesthetic keywords; determining the linkage keywords of each medical aesthetic keyword on each medical aesthetic platform among all other medical aesthetic keywords based on the high-frequency linkage degree; and determining the comprehensive high-frequency linkage degree between each medical aesthetic keyword and its respective linkage keywords on each medical aesthetic platform based on the high-frequency linkage degree between each medical aesthetic keyword and its respective linkage keywords.

[0098] In the embodiment of the present application, first, based on any one medical aesthetic platform, the appearance times (appearance frequency) of each medical aesthetic keyword in the medical aesthetic keyword set are counted according to the medical aesthetic popular data (i.e. the information published in the platform) appeared in the recent period (such as within 2 months), and the appearance times of each medical aesthetic keyword are normalized by using a normalization function (such as maximum-minimum value normalization) to determine the high frequency of each medical aesthetic keyword.

[0099] Secondly, the linkage appearance performance of each medical aesthetic keyword with other medical aesthetic keywords is analyzed, that is, the co-appearance times (frequency) of each medical aesthetic keyword with each of the remaining medical aesthetic keywords in the same information source are traversed, and the co-appearance times are normalized by using a normalization function (such as maximum-minimum value normalization) to determine the linkage appearance coefficient of each medical aesthetic keyword with any one of the remaining medical aesthetic keywords.

[0100] Then, when each medical aesthetic keyword and any one of the remaining medical aesthetic keywords both maintain a relatively high appearance frequency, the higher the linkage appearance coefficient of the two, the more consistent the two are with the extension trend performance of the industry, so the high-frequency linkage degree of each medical aesthetic keyword and any one of the remaining medical aesthetic keywords can be determined. Therefore, the high frequency of each medical aesthetic keyword and any one of the remaining medical aesthetic keywords is averaged, and multiplied by the linkage appearance coefficient of each medical aesthetic keyword and any one of the remaining medical aesthetic keywords (as an adjustment coefficient) to determine the high-frequency linkage degree of each medical aesthetic keyword and any one of the remaining medical aesthetic keywords.

[0101] Then, a high-frequency linkage threshold (e.g., set to 0.85) is set in advance, and the rest of the medical aesthetic keywords with a high-frequency linkage degree greater than or equal to the high-frequency linkage threshold are filtered out from each medical aesthetic keyword and the rest of the medical aesthetic keywords, and these rest of the medical aesthetic keywords are taken as the linkage keywords of each keyword. In this way, the linkage keywords of each medical aesthetic keyword can be obtained.

[0102] Finally, the average value of the high-frequency linkage degrees between each medical aesthetic keyword and its linkage keywords is calculated, and the average value is taken as the comprehensive high-frequency linkage degree between each medical aesthetic keyword and its linkage keywords.

[0103] At this point, the linkage keywords of each medical aesthetic keyword on each medical platform and the comprehensive high-frequency linkage degree between each medical aesthetic keyword and its linkage keywords can be obtained.

[0104] Step S302: Based on the differences between the linkage keywords of each medical aesthetic keyword on different medical platforms and the comprehensive high-frequency linkage degrees between each medical aesthetic keyword and its linkage keywords, the multi-platform medical aesthetic characteristic trend degree of each medical aesthetic keyword is determined.

[0105] Because of the differences in user groups, interaction modes and content dissemination mechanisms between different medical platforms, the popular trends of medical aesthetic related information on each platform may exhibit different characteristics. Therefore, the relative contribution of each medical platform in the popular trend of different medical characteristics needs to be fully considered, so that the comprehensive high-frequency linkage degree between each medical aesthetic keyword and its linkage keywords is corrected based on the relative contribution, and thus the multi-platform medical aesthetic characteristic trend degree of each medical aesthetic keyword is obtained.

[0106] Further, the determination of the multi-platform medical aesthetic characteristic trend degree of each medical aesthetic keyword in step S302 includes: determining the intersection of the linkage keywords of each medical aesthetic keyword on different medical platforms to obtain the multi-platform linkage medical aesthetic keywords of each medical aesthetic keyword; determining the medical aesthetic characteristic trend contribution of the linkage keywords of each medical aesthetic keyword on each medical platform based on the correlation between the linkage keywords of each medical aesthetic keyword on each medical platform and the multi-platform linkage medical aesthetic keywords; and determining the multi-platform medical aesthetic characteristic trend degree of each medical aesthetic keyword based on the comprehensive high-frequency linkage degree and the medical aesthetic characteristic trend contribution of each medical aesthetic keyword on each medical platform.

[0107] In the embodiment of the present application, first, the union of the linkage keywords of each medical aesthetic keyword on different medical platforms is taken as a whole, and the union is recorded as the multi-platform linkage medical aesthetic keywords of each medical aesthetic keyword.

[0108] Secondly, the correlation coefficient between each medical aesthetic keyword and each multi-platform linkage medical aesthetic keyword of each linkage keyword of each medical aesthetic keyword on each medical aesthetic platform is calculated by using a jaccard correlation coefficient, and the correlation coefficient is normalized by using a Softmax normalization, so as to determine the medical aesthetic feature popularity contribution degree of each linkage keyword of each medical aesthetic keyword on each medical aesthetic platform.

[0109] Finally, the medical aesthetic feature popularity contribution degree of each linkage keyword of each medical aesthetic keyword on each medical aesthetic platform is multiplied by the comprehensive high-frequency linkage degree between each medical aesthetic keyword and each linkage keyword of each medical aesthetic platform, and the obtained product is accumulated on multiple medical aesthetic platforms, so as to determine the multi-platform medical aesthetic feature trend degree of each medical aesthetic keyword.

[0110] Step S303: Based on the multi-platform medical aesthetic feature trend degree of each medical aesthetic keyword, and combined with the correlation between all medical aesthetic keywords involved in the medical aesthetic projects recently performed by all medical users in each first medical user group and each multi-platform linkage medical aesthetic keyword of each medical aesthetic keyword, the medical aesthetic feature popularity penetration degree of the medical aesthetic projects recently performed by each first medical user group and each medical aesthetic keyword is determined.

[0111] Through the above process, the medical aesthetic feature trend performance of multiple medical aesthetic platforms is analyzed, and the medical aesthetic user groups divided by the multi-dimensional information in the consumption data of medical aesthetic users. Here, the possible propagation characteristics of individuals in each group can be inferred according to the penetration performance of the medical aesthetic feature trend of multiple medical aesthetic platforms in different user groups in the near future, which helps to more accurately understand the behavior patterns of individuals in the group.

[0112] In the embodiment of the present application, based on the consumption data of medical aesthetic users, the medical aesthetic projects performed by all medical users in each medical aesthetic user group (i.e. the first medical user group) in the near future (such as within 2 months) are obtained. If the medical aesthetic projects performed by these medical users in the near future are more consistent with the popular medical aesthetic features in the near future, it can be inferred that the medical aesthetic user group is more likely to try the related medical aesthetic projects that are popular at the moment in the current and subsequent period.

[0113] In this way, all medical aesthetic keywords involved in the medical aesthetic projects performed by all medical users in each medical aesthetic user group in the near future are used to construct the recent medical aesthetic project keyword set of each medical aesthetic user group. The correlation coefficient between the recent medical aesthetic project keyword set of each medical aesthetic user group and each multi-platform linkage medical aesthetic keyword of each medical aesthetic keyword is calculated by using a jaccard correlation coefficient, and the correlation coefficient is used as the relevance between the recent medical aesthetic project keyword set of each medical aesthetic user group and each multi-platform linkage medical aesthetic keyword of each medical aesthetic keyword.

[0114] Further, the recent medical aesthetic project keyword set of each medical aesthetic user group and the relevance of each medical aesthetic keyword to the multi-platform linked medical aesthetic keyword are taken as weights, and the multi-platform medical aesthetic feature trend degree of each medical aesthetic keyword is weighted and multiplied by the weight, so as to obtain the medical aesthetic feature popular penetration degree of each first medical aesthetic user group.

[0115] Step S304: Statistics of the medical aesthetic feature popular penetration degree of each first medical aesthetic user group and all medical aesthetic keywords, and determination of the medical aesthetic feature popular trend penetration degree of each first medical aesthetic user group.

[0116] In the embodiment of the present application, the medical aesthetic feature popular penetration degree of each medical aesthetic user group is obtained by accumulating the medical aesthetic feature popular penetration degree of each medical aesthetic user group and all medical aesthetic keywords, and normalizing the accumulated value by using a normalization function (such as Sigmoid function).

[0117] Up to now, the medical aesthetic feature popular trend penetration degree of each medical aesthetic user group (medical aesthetic user group) can be obtained.

[0118] Step S400: Based on the consumption data of each medical aesthetic user, and combined with the medical aesthetic popular data of different medical aesthetic platforms, the popular medical aesthetic dynamic characteristics of each medical aesthetic user are analyzed, and combined with the medical aesthetic feature popular trend penetration degree of the first medical aesthetic user group where each medical aesthetic user is located, the medical aesthetic behavior influence degree of each medical aesthetic user to the popular trend is determined.

[0119] The above step S300 analyzes the behavior mode of individuals in the medical aesthetic user group, especially the potential behavior of the medical aesthetic feature popular trend. However, on the basis of group behavior mode, the difference between individuals in the group needs to be concerned, and the popular medical aesthetic dynamic characteristics of each medical aesthetic user need to be analyzed, especially the medical aesthetic technology pursuit degree and individual popular medical aesthetic adaptation degree of each medical aesthetic user. Because these two characteristics belong to the relatively fixed behavior characteristics of medical aesthetic users, they usually do not fluctuate sharply with the change of medical aesthetic popular trend. Therefore, on the basis of group behavior mode, these invariable characteristics are considered, and individual differences are further refined, so as to more accurately predict the behavior response of individuals under a certain popular trend.

[0120] Further, the popular medical aesthetic dynamic characteristics of each medical aesthetic user in the above step S400 include:

[0121] Step S401: Based on the distribution of medical aesthetic projects in the consumption data of each medical aesthetic user, analyze the pursuit of popular medical aesthetic projects by medical aesthetic users, and determine the pursuit degree of popular medical aesthetic technology of each medical aesthetic user.

[0122] The pursuit of popular medical aesthetic technology by the user is essentially the satisfaction of self-image optimization and individualized needs, so the user's extensive degree of medical aesthetic projects in the past can be analyzed, such as the more types of medical aesthetic projects performed and the higher the number of each medical aesthetic project performed, which usually indicates that the user's pursuit degree of popular medical aesthetic technology is higher.

[0123] Further, the determination of the pursuit degree of popular medical aesthetic technology of each medical aesthetic user in the above step S401 includes: determining the actual degree of each medical aesthetic project of each medical aesthetic user based on the number of each medical aesthetic project performed in the consumption data of each medical aesthetic user; determining the real involvement degree of medical aesthetic projects of each medical aesthetic user based on the distribution level of the actual degree of various medical aesthetic projects of each medical aesthetic user; and determining the pursuit degree of popular medical aesthetic technology of each medical aesthetic user based on the number of types of medical aesthetic projects performed in the consumption data of each medical aesthetic user, and combining the real involvement degree of medical aesthetic projects of each medical aesthetic user.

[0124] In the embodiment of the application, the number of types of medical aesthetic projects performed and the number of times of each type performed in the consumption data of each medical aesthetic user are obtained. The number of times of each type of medical aesthetic project of each medical aesthetic user is normalized by using a normalization function (such as maximum-minimum value normalization), so as to obtain the actual degree of each type of medical aesthetic project of each medical aesthetic user. The higher the actual degree of each type of medical aesthetic project of each medical aesthetic user, the higher the input degree of the current user on the project of that type, and the more the user has the actual involvement of the medical aesthetic project of that type.

[0125] Further, the actual degree of each type of medical aesthetic project of each medical aesthetic user is averaged to determine the real involvement degree of medical aesthetic projects of multi-type medical aesthetic projects of each medical aesthetic user. The real involvement degree of medical aesthetic projects of multi-type medical aesthetic projects of each medical aesthetic user is multiplied by the number of types of medical aesthetic projects performed by each medical aesthetic user, and the obtained product value is normalized by using a normalization function (such as Sigmoid function), so as to obtain the pursuit degree of popular medical aesthetic technology of each medical aesthetic user.

[0126] Step S402: Based on the types of medical aesthetic projects performed in the consumption data of each medical aesthetic user, and combined with the recent medical aesthetic popular data of different medical aesthetic platforms, analyze the adaptation of popular medical aesthetic projects to medical aesthetic users, and determine the individual popular medical adaptation degree of each medical aesthetic user.

[0127] Even if the medical beauty user does not know much about the current popular medical beauty project, if the project is highly consistent with the user's own needs, the user's acceptance and trust will be greatly improved. High adaptation means that the project can meet the specific needs of the user, not just popular or hot, so the current popular medical beauty project is more likely to be chosen by the user.

[0128] Further, the individual popular medical adaptation degree of each medical beauty user in the above step S402 includes: constructing an individual medical beauty keyword set of each medical beauty user based on the medical beauty keywords in all medical beauty projects in the consumption data of each medical beauty user; determining the individual popular medical adaptation degree of each medical beauty user based on the multi-platform medical beauty feature trend degree of all medical beauty keywords in the individual medical beauty keyword set.

[0129] In the embodiment of the present application, the individual medical beauty keyword set of each medical beauty user is constructed based on the medical beauty projects performed by the user in the past. Then the multi-platform medical beauty feature trend degree of all medical beauty keywords in the individual medical beauty keyword set is summed up and normalized using a normalization function (such as a Sigmoid function), and the individual popular medical adaptation degree of each medical beauty user is obtained.

[0130] Step S403: Based on the popular medical technology pursuit degree and the individual popular medical adaptation degree, the dynamic characteristics of each medical beauty user are obtained.

[0131] Based on the popular medical technology pursuit degree and the individual popular medical adaptation degree of each medical beauty user determined above, the dynamic characteristics of each medical beauty user are obtained.

[0132] Further, based on the dynamic characteristics of each medical beauty user, the medical beauty behavior influence degree of each medical beauty user to the popular medical beauty trend is analyzed by means of the medical beauty feature trend penetration degree of the group to which each medical beauty user belongs. That is, when the medical beauty feature trend penetration degree of the group to which the medical beauty user belongs is higher, more attention should be paid to the popular medical technology pursuit degree of the medical beauty user; otherwise, more attention should be paid to the individual popular medical adaptation degree of the medical beauty user.

[0133] Further, the medical beauty behavior influence degree of each medical beauty user to the popular trend in the above step S400 includes: determining the weight of the popular medical technology pursuit degree and the individual popular medical adaptation degree of each medical beauty user based on the medical beauty feature trend penetration degree of the first medical beauty user group to which each medical beauty user belongs; and using the weight to add the popular medical technology pursuit degree and the individual popular medical adaptation degree of each medical beauty user to obtain the medical beauty behavior influence degree of each medical beauty user to the popular trend.

[0134] In the embodiment of the present application, based on the dynamic characteristics of each medical beauty user, and combined with the medical beauty feature popular trend penetration of the first medical beauty user group in which each medical beauty user is located, the medical beauty behavior influence degree of each medical beauty user to the popular trend is determined by the following formula :

[0135]

[0136] wherein, represents the medical beauty feature popular trend penetration of the first medical beauty user group in which each medical beauty user is located, and is used as a weight of the popular medical beauty technology pursuit degree; represents the popular medical beauty technology pursuit degree of each medical beauty user; represents the individual popular medical beauty adaptation degree of each medical beauty user, represents the weight of the individual popular medical beauty adaptation degree.

[0137] At this point, the medical beauty behavior influence degree of each medical beauty user to the popular trend can be obtained.

[0138] Step S500: Based on the differences between the medical beauty behavior influence degrees of different medical beauty users, the secondary group division is performed on all medical beauty users, and a plurality of second medical beauty user groups are obtained.

[0139] Through the above process, the medical beauty behavior influence degree of each user to the popular medical beauty feature can be determined, which reflects whether the medical beauty user will have new medical beauty behavior changes in the current group under the influence of the popular medical beauty trend, so that the change of the medical beauty feature label of the user will be brought, that is, the user may choose to try the new medical beauty technology or product under the popular trend.

[0140] In the embodiment of the present application, the medical beauty behavior influence degree of the medical beauty user to the popular medical beauty trend is taken as the fourth dimension feature of the classification method, to re-measure the feature distance between any two medical beauty users, that is, the absolute value of the difference between the medical beauty behavior influence degrees of any two medical beauty users is taken as the feature distance, and based on the feature distance, the DBSCAN clustering method is used again to cluster all medical beauty users, so as to obtain a plurality of medical beauty user groups after secondary classification, and these groups are called second medical beauty user groups.

[0141] Step S600: Based on the medical beauty feature labels corresponding to the first medical beauty user group and the second medical beauty user group in which each medical beauty user is located, the user portrait of each medical beauty user is generated.

[0142] obtain the medical beauty feature label corresponding to each first medical beauty user group and each second medical beauty user group, and generate the user portrait of each medical beauty user based on the medical beauty feature label corresponding to the first medical beauty user group and the second medical beauty user group in which each medical beauty user is located.

[0143] In the embodiment, based on the introduction in the above step S200, the medical beauty feature label corresponding to the first medical beauty user group in which each medical beauty user is located can be obtained. In the same way of obtaining the medical beauty feature label corresponding to the first medical beauty user group in which each medical beauty user is located, a plurality of historical medical beauty user groups and their consumption data can be obtained in advance. In the same way of obtaining the medical beauty behavior influence degree of each medical beauty user, the medical beauty behavior influence degree of each user in the plurality of historical medical beauty user groups obtained in advance can be obtained, and a medical beauty feature label is manually set for each historical medical beauty user group obtained in advance, such as high potential new type, inventory conservative type, etc. Further, by matching the medical beauty behavior influence degree of the medical beauty user in each second medical beauty user group with the medical beauty behavior influence degree of the medical beauty user in each historical medical beauty user group obtained in advance, the most matched historical medical beauty user group of each second medical beauty user group is determined, and the medical beauty feature label of the most matched historical medical beauty user group of each second medical beauty user group is taken as the medical beauty feature label of itself. Since the second feature label of each second medical beauty user group can be obtained.

[0144] At this point, the medical beauty feature label (first feature label and second feature label) corresponding to each first medical beauty user group and each second medical beauty user group can be obtained.

[0145] Further, the medical beauty feature label (first feature label and second feature label) corresponding to the first medical beauty user group and the second medical beauty user group in which each medical beauty user is located is taken as the medical beauty feature label of each medical beauty user, so as to complete the generation of the user portrait of each medical beauty user.

[0146] Based on the same inventive concept, the embodiment of the present application also provides a user portrait generation system for medical beauty users, as shown in Figure 2 The system includes a memory, a processor, and computer program code stored in the memory and running on the processor, wherein when the processor executes the computer program code, the system can execute any of the above-mentioned medical beauty user portrait generation methods.

[0147] Based on the same inventive concept, the embodiment of the present application also provides a user portrait generation device for medical beauty users, as shown in Figure 3 The device includes:

[0148] The data acquisition module is configured to acquire consumption data of different medical beauty users and medical beauty popular data of different medical beauty platforms.

[0149] The first classification module is configured to perform first group division on all the medical beauty users based on differences between the consumption data of different medical beauty users, to obtain a plurality of first medical beauty user groups.

[0150] The penetration analysis module is configured to perform multi-platform medical beauty popular trend analysis based on the medical beauty popular data of different medical beauty platforms and in combination with the consumption data of different medical beauty users in each first medical beauty user group, to determine a medical beauty feature popular trend penetration of each first medical beauty user group.

[0151] The influence analysis module is configured to perform popular medical beauty self-dynamic characteristic analysis on each medical beauty user based on the consumption data of each medical beauty user and in combination with the medical beauty popular data of the different medical beauty platforms, and to determine a medical beauty behavior influence degree of each medical beauty user on a popular trend in combination with the medical beauty feature popular trend penetration of the first medical beauty user group to which each medical beauty user belongs.

[0152] The second classification module is configured to perform second group division on all the medical beauty users based on differences between the medical beauty behavior influence degrees of different medical beauty users, to obtain a plurality of second medical beauty user groups.

[0153] The portrait generation module is configured to generate a user portrait of each medical beauty user based on medical beauty feature labels corresponding to the first medical beauty user group and the second medical beauty user group to which each medical beauty user belongs.

[0154] It should be noted that the apparatus provided in the above embodiments is only exemplarily described based on the division of the above functional modules, and in actual applications, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the computer device is divided into different functional modules to complete all or part of the functions described above.

[0155] The embodiments of the present application can divide the system into functional modules according to the above method examples, for example, each functional module can be provided, or two or more functions can be integrated into one processing module, and the integrated module can be realized in the form of hardware. It should be noted that the division of the modules in the present embodiment is illustrative, and is only a logical function division, and another division mode can be used in actual implementation.

[0156] Based on the same inventive concept, the embodiments of the present application also provide a computer program product, which comprises computer program code, when the computer program code runs on a computer, so that the computer executes any one of the above-mentioned medical beauty user-oriented user portrait generation methods.

[0157] Based on the same inventive concept, the embodiments of the present application also provide a computer readable storage medium storing computer program codes, which, when executed on a computer, cause the computer to perform any of the aforementioned user portrait generation methods for medical and beauty users.

[0158] It should be noted that the above-mentioned embodiments are only used to illustrate the technical solutions of the present application, but not limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for generating user profiles for medical aesthetics users, characterized in that, Includes the following steps: Acquire consumption data from different medical aesthetics users and popular medical aesthetics data from different medical aesthetics platforms; Based on the differences in consumption data among different medical aesthetic users, all medical aesthetic users were first divided into several first medical aesthetic user groups. Based on the medical aesthetics popularity data of different medical aesthetics platforms, and combined with the consumption data of different medical aesthetics users within each first medical aesthetics user group, we conduct multi-platform medical aesthetics popularity trend analysis to determine the penetration of medical aesthetics characteristic popularity trends for each first medical aesthetics user group. Based on the consumption data of each medical aesthetics user and combined with the medical aesthetics popularity data of different medical aesthetics platforms, the dynamic characteristics of popular medical aesthetics are analyzed for each medical aesthetics user. In addition, the penetration degree of the medical aesthetics characteristic popularity trend of each medical aesthetics user's first medical aesthetics user group is combined to determine the degree of influence of popular medical aesthetics behavior on each medical aesthetics user. Based on the differences in the impact of different medical aesthetic behaviors among different medical aesthetic users, all medical aesthetic users are divided into several second medical aesthetic user groups. Based on the medical aesthetic feature tags corresponding to the first and second medical aesthetic user groups to which each medical aesthetic user belongs, a user profile is generated for each medical aesthetic user. Determine the penetration of medical aesthetic characteristics and trends for each primary medical aesthetic user group, including: Based on the recent popular medical aesthetics data of each medical aesthetics platform, the frequency of occurrence of medical aesthetics keywords, and the frequency of co-occurrence of each medical aesthetics keyword with other medical aesthetics keywords in the same information source on each medical aesthetics platform, the linked keywords of each medical aesthetics keyword on each medical aesthetics platform, and the comprehensive high-frequency linkage between each medical aesthetics keyword and its linked keywords are determined. Based on the differences between the linked keywords of each medical aesthetics keyword on different medical aesthetics platforms, and the comprehensive high-frequency linkage between each medical aesthetics keyword and its linked keywords, the popularity trend of each medical aesthetics keyword across multiple platforms is determined. Based on the popularity trend of each medical aesthetics keyword across multiple platforms, and combined with the correlation between all medical aesthetics keywords involved in recent medical aesthetics projects and the multi-platform linked medical aesthetics keywords of each medical aesthetics keyword in the consumption data of all medical aesthetics users in each first medical aesthetics user group, the popularity penetration of medical aesthetics features of each medical aesthetics keyword in the recent medical aesthetics projects of each first medical aesthetics user group is determined. The popularity penetration of medical aesthetic features of each first-tier medical aesthetic user group is determined by statistically analyzing the recent medical aesthetic procedures performed and all medical aesthetic keywords. Determine the associated keywords for each medical aesthetics keyword on each medical aesthetics platform, and the overall high-frequency correlation between each medical aesthetics keyword and its associated keywords, including: Based on the recent popular medical aesthetics data of each medical aesthetics platform, the frequency of occurrence of medical aesthetics keywords is determined; Based on the frequency of each medical aesthetics keyword appearing together with other medical aesthetics keywords in the same information source on each medical aesthetics platform, the correlation coefficient between each medical aesthetics keyword and other medical aesthetics keywords is determined. Based on the frequency and correlation coefficient of each medical aesthetics keyword with other medical aesthetics keywords, the high frequency correlation of each medical aesthetics keyword with other medical aesthetics keywords is determined. Based on the aforementioned high-frequency correlation, the associated keywords for each medical aesthetics keyword on each medical aesthetics platform are determined from all other medical aesthetics keywords. The average value is calculated based on the high-frequency correlation between each medical aesthetics keyword and its associated keywords, and this average value is used as the comprehensive high-frequency correlation between each medical aesthetics keyword and its associated keywords on each medical aesthetics platform. The analysis of the dynamic characteristics of popular cosmetic procedures for each user includes: Based on the consumption data of each medical aesthetics user, the distribution of medical aesthetics projects is analyzed, the pursuit of popular medical aesthetics projects by medical aesthetics users is analyzed, and the degree of pursuit of popular medical aesthetics techniques by each medical aesthetics user is determined. Based on each medical aesthetics user's consumption data, the types of medical aesthetics projects are identified, and combined with recent medical aesthetics trend data from different medical aesthetics platforms, the suitability of popular medical aesthetics projects for medical aesthetics users is analyzed to determine the individual suitability of each medical aesthetics user for popular medical aesthetics projects. Based on the popularity of certain cosmetic medical technologies and the individual's suitability for those technologies, the dynamic characteristics of each cosmetic medical user are obtained.

2. The user profile generation method for medical aesthetics users according to claim 1, characterized in that, Determining the popularity trend of each medical aesthetics keyword across multiple platforms includes: Determine the intersection of each medical aesthetics keyword with its associated keywords on different medical aesthetics platforms to obtain the multi-platform associated medical aesthetics keywords for each medical aesthetics keyword; Based on the correlation between each linked keyword of each medical aesthetics keyword on each medical aesthetics platform and the linked medical aesthetics keywords of each multi-platform, the contribution of each linked keyword of each medical aesthetics keyword on each medical aesthetics platform to the popularity of medical aesthetics features is determined. Based on the comprehensive high-frequency correlation and the contribution of the popularity of each medical aesthetics keyword on various medical aesthetics platforms, the popularity trend of each medical aesthetics keyword across multiple platforms is determined.

3. The user profile generation method for medical aesthetics users according to claim 1, characterized in that, Determine the level of interest in popular cosmetic techniques among each cosmetic patient, including: Based on each medical aesthetic user's consumption data, the number of times each medical aesthetic procedure was performed was determined to assess the actual degree of each medical aesthetic procedure performed by each user. Based on the distribution level of the actual number of cosmetic procedures performed by each cosmetic user, the actual number of cosmetic procedures involved by each user is determined. Based on the consumption data of each medical aesthetics user, the number of types of medical aesthetics projects is counted, and combined with the actual involvement of each medical aesthetics user in medical aesthetics projects, the degree of pursuit of popular medical aesthetics techniques by each medical aesthetics user is determined.

4. The user profile generation method for medical aesthetics users according to claim 1, characterized in that, Determine the individual suitability of each cosmetic surgery user for popular cosmetic procedures, including: Based on each medical aesthetic user's consumption data, medical aesthetic keywords are extracted from all medical aesthetic projects to construct an individual medical aesthetic keyword set for each medical aesthetic user; Based on the popularity trend of all medical aesthetic keywords in the individual medical aesthetic keyword set across multiple platforms, the individual popularity medical aesthetic suitability of each medical aesthetic user is determined.

5. The user profile generation method for medical aesthetics users according to claim 1, characterized in that, Determine the degree to which each cosmetic surgery user is influenced by popular trends in cosmetic surgery behavior, including: Based on the penetration of the medical aesthetics characteristics and trends of the primary medical aesthetics user group to which each medical aesthetics user belongs, the weights of each medical aesthetics user's pursuit of popular medical aesthetics techniques and individual suitability for popular medical aesthetics are determined. Using the aforementioned weights, the degree of pursuit of popular cosmetic techniques and the individual suitability of each cosmetic user are weighted and summed to obtain the degree to which each cosmetic user's cosmetic behavior is affected by popular trends.

6. The user profile generation method for medical aesthetics users according to claim 1, characterized in that, The initial segmentation of different cosmetic surgery users yielded several primary cosmetic surgery user groups, including: Based on the difference in the consumption amount of medical aesthetic projects between any two medical aesthetic users, determine the first difference value; Based on the differences in frequency values ​​between each type of medical aesthetic procedure in the consumption data of any individual medical aesthetic user, a second difference value is determined. Based on the consumption data of any individual medical aesthetic user, the differences between the medical aesthetic care areas of medical aesthetic projects are analyzed to determine the third difference value; Based on the first difference value, the second difference value, and the third difference value, determine the multi-dimensional feature distance between any two medical aesthetic users; Based on the multi-dimensional feature distance, a clustering algorithm is used to cluster all medical aesthetic users, and the resulting clusters are used as several first medical aesthetic user groups.

7. A user profile generation system for medical aesthetics users, characterized in that, The device includes a memory, a processor, and executable computer program code stored in the memory and executable on the processor. When the processor executes the computer program code, it performs a user profile generation method for medical aesthetics users as described in any one of claims 1 to 6.

Images