Clothing matching intelligent recommendation platform based on cloud computing

Abstract

The invention discloses an intelligent garment matching recommendation platform based on cloud computing, and relates to the technical field of garment matching recommendation, and the platform comprises a first obtaining module which obtains a garment type needing to be recommended according to a first garment needing to be matched by a first user; the second acquisition module is used for acquiring a second garment with the maximum matching score with the first garment in the cloud database according to a preset garment matching algorithm; the third acquisition module is used for acquiring third clothes which are matched with the second clothes in the cloud database and accord with the clothes type needing to be recommended; and the intelligent recommendation module is used for extracting features of the third clothes, comprehensively scoring the third clothes according to the features, and carrying out descending sort recommendation according to the comprehensive scores. Through cloud computing and multi-module cooperative work, the method can accurately analyze the demands of the user, thereby providing highly personalized clothing matching recommendation, improving the recommendation accuracy, and improving the satisfaction of the user.

Description

Technical Field

[0001] The present invention relates to the technical field of clothing matching recommendation, and in particular to a clothing matching intelligent recommendation platform based on cloud computing. Background Art

[0002] In the existing technology, clothing matching recommendations already exist, and recommendations are made through simple rules and basic user information (such as height, weight, gender, etc.). However, the recommendation results are often not accurate enough and cannot fully meet the personalized needs of users. Especially in the face of diverse and complex user needs, the recommendation accuracy and user satisfaction of existing systems need to be improved. The shortcomings of existing technologies include: the singleness of the recommendation algorithm and the lack of multi-dimensional data analysis capabilities; the low degree of personalized recommendations and the inability to fully consider the user's behavioral characteristics and preferences; the lack of a dynamic adjustment mechanism for user feedback, resulting in insufficient timeliness and accuracy of recommendation results; the limited ability to process and analyze user big data, and the inability to provide accurate and real-time recommendation services. Summary of the invention

[0003] Based on the above-mentioned shortcomings of the prior art, the purpose of the present invention is to provide a clothing matching intelligent recommendation platform based on cloud computing to solve the above-mentioned technical problems.

[0004] To achieve the above purpose, the present invention provides the following technical solution: a cloud computing-based clothing matching intelligent recommendation platform, comprising:

[0005] A first acquisition module: acquiring the type of clothing to be recommended according to the first clothing that the first user needs to match;

[0006] A second acquisition module: acquiring a second garment having the largest matching score with the first garment in the cloud database according to a preset garment matching algorithm;

[0007] A third acquisition module: acquiring a third garment in the cloud database that has been matched with the second garment and meets the type of garment to be recommended;

[0008] Intelligent recommendation module: extract the features of the third clothing, comprehensively score the third clothing according to the features, and recommend them in descending order according to the comprehensive scores.

[0009] The present invention is further configured such that the logic for obtaining the first clothing that the first user needs to match includes:

[0010] Choose the clothing that exists;

[0011] Or calculate the user similarity score according to the user portrait of the first user and a preset user portrait matching algorithm, obtain a similar user table in descending order of the user similarity scores, obtain a fourth garment that meets the screening criteria according to the similar user table, and display the fourth garment for selection.

[0012] The present invention is further configured that the user portrait includes basic information and behavioral characteristics, wherein the basic information includes height, weight and gender, and the behavioral characteristics include browsing, clicking and purchasing; and calculating the user similarity score according to the user portrait and a preset user portrait matching algorithm, including:

[0013] Perform screening and matching based on the basic information of the first user, and select multiple second users with similar basic information;

[0014] Obtaining behavior characteristics of multiple second users, and calculating browsing similarity, click similarity, and purchase similarity, as well as browsing similarity frequency, click similarity frequency, and purchase similarity frequency according to the behavior characteristics of the first user;

[0015] The browsing similarity frequency, the click similarity frequency, and the purchase similarity frequency are normalized, and the user similarity score is calculated based on the browsing similarity, the click similarity, the purchase similarity, and the normalized browsing similarity frequency, the click similarity frequency, and the purchase similarity frequency.

[0016] The present invention is further configured such that the calculation logic of the browsing similarity, the click similarity and the purchase similarity is: Among them, the values ​​of x are b, c and p, S b , S c and S p are respectively browsing similarity, clicking similarity and purchasing similarity, b1, c1 and p1 are respectively the clothing sets browsed, clicked and purchased by the first user, b2, c2 and p2 are respectively the clothing sets browsed, clicked and purchased by the second user; the calculation logic of the browsing similarity frequency, clicking similarity frequency and purchasing similarity frequency is: Among them, f b 、f c and f p They are browsing similarity frequency, clicking similarity frequency and purchasing similarity frequency.

[0017] The present invention is further configured to normalize the browsing similarity frequency, the clicking similarity frequency and the purchasing similarity frequency, and the calculation logic is: in, and They represent the normalized browsing similarity frequency, click similarity frequency, and purchase similarity frequency respectively; the calculation logic of user similarity score is: Among them, Stotal Rate user similarity.

[0018] The present invention is further configured to obtain the second garment having the largest matching score with the first garment in the cloud database according to a preset garment matching algorithm, including:

[0019] Calculate the third-order moment of each color channel in the clothing image, and set the third-order moment as a color feature;

[0020] Calculating the grayscale values ​​of each pixel and surrounding pixels in the clothing image to generate a local binary pattern, and setting the local binary pattern as a texture feature;

[0021] Calculating the gradient direction and magnitude of edge pixels in the image, obtaining an edge gradient direction histogram, and setting the edge gradient direction histogram as a shape feature;

[0022] The cosine similarities of the color feature, the texture feature, and the shape feature are calculated and weighted to obtain a matching score.

[0023] The present invention is further configured to extract the features of the third clothing and comprehensively score the third clothing according to the features, including:

[0024] Calculate the color difference between the second garment and the third garment according to the Lab color space;

[0025] Calculate the style similarity of the colors of the second garment and the third garment according to the neural network model;

[0026] A comprehensive score is calculated based on the color difference and the style similarity.

[0027] The present invention is further configured such that the calculation logic of the color difference is: Among them, L2 and L3 represent the brightness components of the second clothing and the third clothing in the Lab color space, a2 and a3 represent the green to red components of the second clothing and the third clothing in the Lab color space, and b2 and b3 represent the blue to yellow components of the second clothing and the third clothing in the Lab color space.

[0028] The present invention is further configured to calculate the style similarity of the second clothing and the third clothing colors according to the neural network model, including:

[0029] Inputting a second clothing image and a third clothing image that have been preprocessed, wherein the preprocessing includes resizing and pixel normalization;

[0030] The convolution layer uses the convolution kernel to scan the image and extract low-level features. The pooling layer reduces the size of the feature map to reduce the amount of calculation and overfitting.

[0031] The output of the previous convolutional layer is flattened through a fully connected layer and linearly transformed through a set of weights and biases to output a style feature vector;

[0032] The cosine similarity of the style feature vectors of the second clothing image and the third clothing image is calculated to obtain the style similarity of the second clothing image and the third clothing image.

[0033] The present invention is further configured such that the calculation logic of the comprehensive score is: Among them, CS total is the comprehensive score, S s is the style similarity.

[0034] The present invention provides a cloud computing-based clothing matching intelligent recommendation platform, the platform comprising a first acquisition module: acquiring a type of clothing to be recommended according to a first clothing that a first user needs to match; a second acquisition module: acquiring a second clothing with the largest matching score with the first clothing in a cloud database according to a preset clothing matching algorithm; a third acquisition module: acquiring a third clothing in the cloud database that has been matched with the second clothing and meets the type of clothing to be recommended; an intelligent recommendation module: extracting features of the third clothing, comprehensively scoring the third clothing according to the features, and recommending the third clothing in descending order according to the comprehensive score, the beneficial effects produced include:

[0035] 1. Improve the accuracy of recommendations: Through cloud computing and multi-module collaboration, the present invention can accurately analyze user needs and provide highly personalized clothing matching recommendations, which not only improves the accuracy of recommendations, but also improves user satisfaction;

[0036] 2. Multi-dimensional data analysis: The present invention uses the basic information and behavioral characteristics of users to perform multi-dimensional data processing and analysis. Through comprehensive analysis of user behavioral characteristics such as click frequency, purchase frequency, browsing habits, etc., it can achieve more comprehensive and in-depth user demand mining and improve recommendation effects;

[0037] 3. Enhance user experience: The present invention provides personalized clothing recommendations, making it more convenient and satisfying for users to choose clothing combinations, thereby improving user experience.

[0038] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0040] Figure 1 The schematic diagram of the structure of a clothing matching intelligent recommendation platform based on cloud computing is shown as an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0041] The following will describe the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention, not for limiting the scope of protection of the present invention.

[0042] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0043] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it is obvious to those skilled in the art that the embodiments of the present invention can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present invention difficult to understand.

[0044] Cloud computing-based clothing matching intelligent recommendation platform, such as Figure 1 As shown, including:

[0045] A first acquisition module: acquiring the type of clothing to be recommended according to the first clothing that the first user needs to match;

[0046] A second acquisition module: acquiring a second garment having the largest matching score with the first garment in the cloud database according to a preset garment matching algorithm;

[0047] A third acquisition module: acquiring a third garment in the cloud database that has been matched with the second garment and meets the type of garment to be recommended;

[0048] Intelligent recommendation module: extract the features of the third clothing, comprehensively score the third clothing according to the features, and recommend them in descending order according to the comprehensive scores.

[0049] Specifically, the first clothing that the first user needs to match refers to the clothing that the user selects and needs to match, and this clothing is the starting point of the entire matching recommendation process; the present invention is further configured that the platform will recommend other suitable clothing according to the characteristics of this clothing, and the acquisition logic of the first clothing that the first user needs to match includes:

[0050] Select existing clothing; specifically, the user can directly select existing clothing as the first clothing on the platform, which may be clothing that the user has already purchased or collected;

[0051] Or calculate the user similarity score according to the user portrait of the first user and a preset user portrait matching algorithm, obtain a similar user table in descending order of the user similarity scores, obtain the fourth clothing that meets the screening criteria according to the similar user table, and display the fourth clothing for selection. Specifically, when the user needs the platform to make recommendations, the platform recommends new clothing based on the choices of similar users. Through the above logic, not only the user experience is improved, but also the personalization and diversity of the recommendations are enhanced, and the accuracy and satisfaction of the recommendations are improved.

[0052] The present invention is further configured that the user portrait includes basic information and behavioral characteristics, wherein the basic information includes height, weight and gender, and the behavioral characteristics include browsing, clicking and purchasing; and calculating the user similarity score according to the user portrait and a preset user portrait matching algorithm, including:

[0053] Screening and matching are performed based on the basic information of the first user, and multiple second users with similar basic information are selected; by matching the basic information, it is ensured that the recommended clothing is more in line with the basic needs of the user, and the reliability of the recommendation is increased; here, screening and matching are performed based on the basic information of the first user, that is, users with similar height and weight and the same gender as the first user are selected as second users, and the similarity here can be set to an interval, and there is no restriction on this here.

[0054] The behavior characteristics of the plurality of second users are obtained, and the browsing similarity, the click similarity and the purchase similarity, as well as the browsing similarity frequency, the click similarity frequency and the purchase similarity frequency are calculated according to the behavior characteristics of the first user; the present invention is further configured such that the calculation logic of the browsing similarity, the click similarity and the purchase similarity is: Among them, the values ​​of x are b, c and p, S b , S c and S pare respectively browsing similarity, click similarity and purchase similarity, b1, c1 and p1 are respectively the clothing collections browsed, clicked and purchased by the first user, b2, c2 and p2 are respectively the clothing collections browsed, clicked and purchased by the second user. By calculating the behavioral characteristics, we can more comprehensively understand the interests and preferences of users, so as to make more accurate recommendations; the calculation logic of the browsing similarity frequency, the click similarity frequency and the purchase similarity frequency is: Among them, f b 、f c and f p They are browsing similarity frequency, clicking similarity frequency and purchasing similarity frequency.

[0055] The browsing similarity frequency, the click similarity frequency and the purchase similarity frequency are normalized, and the user similarity score is calculated according to the browsing similarity, the click similarity and the purchase similarity, and the normalized browsing similarity frequency, the click similarity frequency and the purchase similarity frequency. The present invention is further configured to normalize the browsing similarity frequency, the click similarity frequency and the purchase similarity frequency, and the calculation logic is: in, and They represent the normalized browsing similarity frequency, click similarity frequency, and purchase similarity frequency respectively. Through normalization, the features of different frequencies can be compared on the same scale, eliminating the influence of dimension and ensuring the fairness of similarity calculation. The calculation logic of user similarity score is: Among them, S total By constructing user portraits and calculating similarity logic, the platform can provide more accurate and personalized clothing recommendations. It takes into account the user's basic information and behavioral characteristics, calculates the user's similarity score, and thus selects clothing that matches the user's needs, thereby improving the efficiency of the recommendation system and user experience.

[0056] The present invention is further configured to obtain the second garment having the largest matching score with the first garment in the cloud database according to a preset garment matching algorithm, including:

[0057] Calculate the third-order moment of each color channel in the clothing image, and set the third-order moment as the color feature. Specifically, the color information of the image can be represented by the third-order moment of the color channel. The third-order moment is a statistic of the color feature of the image, which reflects the shape and symmetry of the color distribution. By calculating the third-order moment, the color feature of the clothing image can be extracted, the accuracy and completeness of the color feature are ensured, and color matching is helpful.

[0058] Calculate the gray value of each pixel in the clothing image and the surrounding pixels to generate a local binary pattern, and set the local binary pattern as a texture feature. Specifically, a local binary pattern (LBP) is a method for describing the local texture features of an image. By calculating the gray value of each pixel and comparing it with the surrounding pixels, a binary pattern is generated. Through LBP, the texture features of the clothing image can be extracted, which helps to capture the detailed information of the image and enhance the accuracy of texture matching;

[0059] Calculate the gradient direction and magnitude of edge pixels in the image, obtain the edge gradient direction histogram, and set the edge gradient direction histogram as the shape feature. Specifically, the edge gradient direction and magnitude can reflect the shape features in the image. By calculating the gradient direction and magnitude of edge pixels and generating a gradient direction histogram, the shape information of the image is captured. Through the edge gradient direction histogram, the shape features of the clothing image can be extracted to ensure the accuracy and completeness of shape matching;

[0060] The cosine similarities of the color feature, the texture feature, and the shape feature are calculated and weighted to obtain a matching score. Specifically, through the above logic, the platform can accurately obtain the second clothing with the highest matching score with the first clothing selected by the user from the cloud database.

[0061] The present invention is further configured to extract the features of the third clothing and comprehensively score the third clothing according to the features, including:

[0062] The color difference between the second garment and the third garment is calculated according to the Lab color space; the present invention is further configured such that the calculation logic of the color difference is: Among them, L2 and L3 represent the brightness components of the second garment and the third garment in the Lab color space, a2 and a3 represent the green to red components of the second garment and the third garment in the Lab color space, b2 and b3 represent the blue to yellow components of the second garment and the third garment in the Lab color space. Specifically, the Lab color space is a color space based on human visual perception, which includes three components: L (brightness), a (green to red), and b (blue to yellow). Through the above color difference calculation formula, the platform can more accurately evaluate the color difference between the second garment and the third garment, thereby recommending color-coordinated clothing combinations.

[0063] The style similarity of the second clothing and the third clothing colors is calculated according to the neural network model; the present invention is further configured to calculate the style similarity of the second clothing and the third clothing colors according to the neural network model, including:

[0064] Inputting the preprocessed second clothing image and the third clothing image, wherein the preprocessing includes resizing and pixel normalization; specifically, by resizing and normalizing the pixels, the images are converted into a unified format to facilitate neural network processing;

[0065] The convolution layer uses the convolution kernel to scan the image and extract low-level features. The pooling layer reduces the size of the feature map to reduce the amount of calculation and overfitting. Furthermore, the main operation of the convolution layer is convolution calculation, and the calculation logic is: Where I is the input image or input feature map, K is the convolution kernel, (x, y) is the position of the output feature map, and m and n are the sizes of the convolution kernel. The pooling layer is used to reduce the size of the feature map. The calculation logic of the maximum pooling is: Among them, P is the feature map after pooling, F is the input feature map, and kernel is the pooling window.

[0066] The fully connected layer flattens the output of the previous convolutional layer and performs a linear transformation through a set of weights and biases to output a style feature vector. Specifically, the fully connected layer flattens the output of the previous layer and performs a linear transformation through a set of weights and biases. Its calculation logic is: z=W*x+b, where z is the style feature vector, W is the weight matrix, x is the output of the previous convolutional layer, and b is the bias vector.

[0067] The cosine similarity of the style feature vectors of the second clothing image and the third clothing image is calculated to obtain the style similarity of the second clothing image and the third clothing image. Through the above logic, the style similarity of the second clothing and the third clothing color is calculated based on the neural network model. The system can accurately extract the style features of the clothing image and calculate the similarity, thereby improving the accuracy of the recommendation and user satisfaction. This method not only improves the accuracy of style matching, but also enhances the robustness and computing performance of the recommendation system, providing users with better clothing matching suggestions. Specifically, through the above logic, the style similarity of the second clothing and the third clothing color is calculated based on the neural network model, which can accurately extract the style features of the clothing image and calculate the similarity, thereby improving the accuracy of the recommendation and user satisfaction. This method not only improves the accuracy of style matching, but also enhances the robustness and computing performance of the recommendation system, thereby providing users with better clothing matching suggestions.

[0068] A comprehensive score is calculated according to the color difference and the style similarity; the present invention is further configured such that the calculation logic of the comprehensive score is: the calculation logic of the comprehensive score is: Among them, CS total is the comprehensive score, S sFor style similarity, through the above logic, we can comprehensively consider color and style characteristics and calculate the comprehensive score of clothing combinations, which can provide users with more accurate and personalized clothing matching recommendations and improve the quality of recommendation results and user experience.

[0069] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0070] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0071] In this application, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0072] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0073] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0074] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0075] In the several embodiments provided in the present application, it should be understood that the disclosed system can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0076] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0077] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0078] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.

[0079] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. The cloud computing-based clothing matching intelligent recommendation platform is characterized by: include: A first acquisition module: acquiring the type of clothing to be recommended according to the first clothing that the first user needs to match; A second acquisition module: acquiring a second garment having the largest matching score with the first garment in the cloud database according to a preset garment matching algorithm; A third acquisition module: acquiring a third garment in the cloud database that has been matched with the second garment and meets the type of garment to be recommended; Intelligent recommendation module: extract the features of the third clothing, comprehensively score the third clothing according to the features, and recommend them in descending order according to the comprehensive scores.

2. The cloud computing-based clothing matching intelligent recommendation platform according to claim 1 is characterized in that: The logic for obtaining the first clothing that the first user needs to match includes: Choose the clothing that exists; Or calculate the user similarity score according to the user portrait of the first user and a preset user portrait matching algorithm, obtain a similar user table in descending order of the user similarity scores, obtain a fourth garment that meets the screening criteria according to the similar user table, and display the fourth garment for selection.

3. The cloud computing-based clothing matching intelligent recommendation platform according to claim 2 is characterized in that: The user portrait includes basic information and behavioral characteristics, where the basic information includes height, weight and gender, and the behavioral characteristics include browsing, clicking and purchasing; the user similarity score is calculated based on the user portrait and the preset user portrait matching algorithm, including: Perform screening and matching based on the basic information of the first user, and select multiple second users with similar basic information; Obtaining behavior characteristics of multiple second users, and calculating browsing similarity, click similarity, and purchase similarity, as well as browsing similarity frequency, click similarity frequency, and purchase similarity frequency according to the behavior characteristics of the first user; The browsing similarity frequency, the click similarity frequency, and the purchase similarity frequency are normalized, and the user similarity score is calculated based on the browsing similarity, the click similarity, the purchase similarity, and the normalized browsing similarity frequency, the click similarity frequency, and the purchase similarity frequency.

4. The cloud computing-based clothing matching intelligent recommendation platform according to claim 3 is characterized in that: The calculation logic of the browsing similarity, click similarity and purchase similarity is: Among them, the values ​​of x are b, c and p, S b , S c and S p are respectively browsing similarity, clicking similarity and purchasing similarity, b1, c1 and p1 are respectively the clothing sets browsed, clicked and purchased by the first user, b2, c2 and p2 are respectively the clothing sets browsed, clicked and purchased by the second user; the calculation logic of the browsing similarity frequency, clicking similarity frequency and purchasing similarity frequency is: Among them, f b 、f c and f p They are browsing similarity frequency, clicking similarity frequency and purchasing similarity frequency.

5. The cloud computing-based clothing matching intelligent recommendation platform according to claim 4 is characterized in that: Normalize the browsing similarity frequency, click similarity frequency, and purchase similarity frequency. The calculation logic is: in, and They represent the normalized browsing similarity frequency, click similarity frequency, and purchase similarity frequency respectively; the calculation logic of user similarity score is: Among them, S total Rate user similarity.

6. The cloud computing-based clothing matching intelligent recommendation platform according to claim 1, characterized in that: Acquiring the second garment having the largest matching score with the first garment in the cloud database according to a preset garment matching algorithm includes: Calculate the third-order moment of each color channel in the clothing image, and set the third-order moment as a color feature; Calculating the grayscale values ​​of each pixel and surrounding pixels in the clothing image to generate a local binary pattern, and setting the local binary pattern as a texture feature; Calculating the gradient direction and magnitude of edge pixels in the image, obtaining an edge gradient direction histogram, and setting the edge gradient direction histogram as a shape feature; The cosine similarities of the color feature, the texture feature, and the shape feature are calculated and weighted to obtain a matching score.

7. The cloud computing-based clothing matching intelligent recommendation platform according to claim 1, characterized in that: Extracting the features of the third clothing and comprehensively scoring the third clothing according to the features, including: Calculate the color difference between the second garment and the third garment according to the Lab color space; Calculate the style similarity of the colors of the second garment and the third garment according to the neural network model; A comprehensive score is calculated based on the color difference and the style similarity.

8. The cloud computing-based clothing matching intelligent recommendation platform according to claim 7 is characterized in that: The calculation logic of color difference is: Among them, L2 and L3 represent the brightness components of the second clothing and the third clothing in the Lab color space, a2 and a3 represent the green to red components of the second clothing and the third clothing in the Lab color space, and b2 and b3 represent the blue to yellow components of the second clothing and the third clothing in the Lab color space.

9. The cloud computing-based clothing matching intelligent recommendation platform according to claim 8, characterized in that: The style similarity between the colors of the second garment and the third garment is calculated according to the neural network model, including: Inputting a second clothing image and a third clothing image that have been preprocessed, wherein the preprocessing includes resizing and pixel normalization; The convolution layer uses the convolution kernel to scan the image and extract low-level features. The pooling layer reduces the size of the feature map to reduce the amount of calculation and overfitting. The output of the previous convolutional layer is flattened through a fully connected layer and linearly transformed through a set of weights and biases to output a style feature vector; The cosine similarity of the style feature vectors of the second clothing image and the third clothing image is calculated to obtain the style similarity of the second clothing image and the third clothing image.

10. The cloud computing-based clothing matching intelligent recommendation platform according to claim 9, characterized in that: The calculation logic of the comprehensive score is: Among them, CS total is the comprehensive score, S s is the style similarity.

Images