Service recommendation method and device based on user behaviors, equipment and medium

Abstract

The invention relates to the technical field of data analysis, can be applied to business system platforms of financial science and technology, medical health and the like, and discloses a service recommendation method, device, equipment and medium based on user behaviors, the method comprises the following steps: obtaining a to-be-marked area of a platform, carrying out data point burying on the to-be-marked area of the platform to obtain a point burying area, tracking browsing information of a preset target user in the burying point area in real time to obtain user behavior data, generating a plurality of user behavior paths according to the user behavior data, obtaining historical behavior data of the preset target user, and performing preference analysis on the preset target user according to the historical behavior data and the user behavior paths to obtain a preference analysis result; and obtaining service preference information, obtaining a platform browsing area corresponding to the user behavior path, and pushing the service preference information to the platform browsing area of a preset target user. According to the method, the utilization rate of the user behavior path and the accuracy of recommending personalized contents can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and in particular to a method, device, equipment and medium for service recommendation based on user behavior. Background Art

[0002] With the continuous advancement of technology, most of the existing platforms already support intelligent learning and decision-making. When buyers communicate with merchants and request returns or exchanges due to product problems, the platform will intelligently determine that it is the merchant or seller's responsibility. The platform-supported processing entrance will automatically pop up in the chat record, and the order will initiate a return or exchange. In the absence of freight insurance, the buyer's responsibility will be automatically returned to the buyer, and the freight will be claimed.

[0003] However, the existing Ping An Good Car Owner platform does not track all user behavior paths and only performs basic analysis of users' current behavior. This makes it difficult for the platform to accurately recommend personalized content or products and cannot provide users with a tailored experience. The lack of comprehensive data collection and analysis means that the platform lacks data support when making operational decisions, and is unable to promptly identify changes in user needs or optimize potential pain points. The platform's marketing strategies often cannot be accurately delivered to target users, reducing the conversion rate of marketing investment and wasting resources.

[0004] In the healthcare sector, users can be provided with customized treatment plans, medication recommendations, expert advice, and other services based on their health records and behavioral data. For example, when a user searches for symptoms of a particular disease, the platform can recommend relevant doctors or hospitals based on their browsing history and arrange online appointments. Without analyzing health data across all user behavior paths, potential health issues may not be predicted, resulting in lower accuracy in the personalized health advice provided to users.

[0005] In the fintech sector, financial platforms can understand user needs and interests by analyzing user behavior data such as visits, purchases, and interactions on their platforms. Based on user behavior, financial institutions can implement targeted marketing strategies, such as promoting customized financial products, loan offers, or credit card promotions, to increase user engagement and loyalty. Without analyzing all user behavior paths, it may be difficult to provide users with the products they truly desire.

[0006] Therefore, current technologies have the problem of low utilization of user behavior paths and difficulty in accurately recommending personalized content. Summary of the Invention

[0007] The present invention provides a service recommendation method, apparatus, device and medium based on user behavior, the main purpose of which is to solve the problems of low utilization of user behavior paths and difficulty in accurately recommending personalized content.

[0008] In a first aspect, to achieve the above-mentioned objectives, the present invention provides a service recommendation method based on user behavior, comprising:

[0009] Obtaining the area to be marked on the platform, performing data embedding on the area to be marked, and obtaining the embedding area;

[0010] Real-time tracking of the browsing information of the preset target user in the tracking area to obtain user behavior data;

[0011] generating a plurality of user behavior paths according to the user behavior data;

[0012] Obtaining historical behavior data of a preset target user, performing preference analysis on the preset target user based on the historical behavior data and the user behavior path, and obtaining service preference information;

[0013] The platform browsing area corresponding to the user behavior path is obtained, and the service preference information is pushed to the platform browsing area of ​​the preset target user.

[0014] In a second aspect, the present invention further provides a service recommendation device based on user behavior, comprising:

[0015] The data embedding module is used to obtain the area to be marked on the platform, embed data in the area to be marked, and obtain the embedding area;

[0016] A data tracking module is used to track the browsing information of preset target users in the tracking area in real time to obtain user behavior data;

[0017] A path generation module, configured to generate a plurality of user behavior paths based on the user behavior data;

[0018] A preference analysis module is used to obtain historical behavior data of a preset target user, perform preference analysis on the preset target user based on the historical behavior data and the user behavior path, and obtain service preference information;

[0019] The service recommendation module is used to obtain the platform browsing area corresponding to the user behavior path and push the service preference information to the platform browsing area of ​​the preset target user.

[0020] In a third aspect, the present invention further provides an electronic device, comprising:

[0021] at least one processor; and,

[0022] a memory communicatively connected to the at least one processor; wherein,

[0023] The memory stores a computer program that can be executed by the at least one processor. The computer program is executed by the at least one processor so that the at least one processor can execute the above-mentioned service recommendation method based on user behavior.

[0024] In a fourth aspect, the present invention also provides a computer-readable storage medium, in which at least one computer program is stored. The at least one computer program is executed by a processor in an electronic device to implement the above-mentioned user behavior-based service recommendation method.

[0025] The present invention obtains the area to be marked of the platform, performs data embedding on the area to be marked, and obtains the embedding area. By embedding data on the area to be marked of the platform and obtaining embedding marking parameters, the user's behavior and interaction on the platform can be accurately captured, and the browsing information of the preset target user in the embedding area is tracked in real time to obtain user behavior data. Based on these behavior data, the platform can more accurately predict the user's needs, generate a number of user behavior paths according to the user behavior data, and clearly understand the user's interaction process and behavior sequence within a specific time period, obtain the historical behavior data of the preset target user, perform preference analysis on the preset target user according to the historical behavior data and the user behavior path, and obtain service preference information, accurately capture the user's needs and preferences, obtain the platform browsing area corresponding to the user behavior path, and push the service preference information to the platform browsing area of ​​the preset target user, which can effectively improve the utilization rate of the user behavior path and the accuracy of recommending personalized content. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments of the present invention. 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 paying any creative labor.

[0027] Figure 1 A schematic diagram of an application environment for a service recommendation method based on user behavior in one embodiment of the present invention;

[0028] Figure 2 A flowchart of a method for recommending services based on user behavior according to an embodiment of the present invention;

[0029] Figure 3 A schematic diagram of a process for analyzing user preferences in a method for recommending services based on user behavior according to an embodiment of the present invention;

[0030] Figure 4 A schematic diagram of modules of a service recommendation device based on user behavior provided by one embodiment of the present invention;

[0031] Figure 5 A schematic structural diagram of an electronic device for implementing a method for recommending services based on user behavior according to an embodiment of the present invention;

[0032] Figure 6 Another structural diagram of an electronic device for implementing a service recommendation method based on user behavior provided by an embodiment of the present invention.

[0033] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0034] In order to enable those skilled in the art to better understand the technical solutions of the present disclosure, and to fully understand and implement how the present disclosure applies technical means to solve technical problems and achieve the corresponding technical effects, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. The embodiments of the present disclosure and the various features in the embodiments can be combined with each other without conflict, and the technical solutions formed are all within the scope of protection of the present disclosure. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present disclosure.

[0035] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, apparatus, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0036] An embodiment of the present application provides a service recommendation method based on user behavior, and the execution subject of the service recommendation method based on user behavior includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the device provided by the embodiment of the present application. In other words, the service recommendation method based on user behavior can be executed by software or hardware installed on a terminal device or a server device. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0037] The present invention provides a service recommendation method based on user behavior, which can be applied in the following situations: Figure 1 application environment. Among them, the client communicates with the server through the network. The server can obtain the area to be marked of the platform through the client, perform data burying on the area to be marked, and obtain the burying area. By burying data on the area to be marked of the platform and obtaining the burying marking parameters, the user's behavior and interaction on the platform can be accurately captured, and the browsing information of the preset target user in the burying area can be tracked in real time to obtain user behavior data. Based on these behavior data, the platform can more accurately predict the user's needs, generate several user behavior paths according to the user behavior data, and clearly understand the user's interaction process and behavior sequence in a specific time period, obtain the historical behavior data of the preset target user, perform preference analysis on the preset target user according to the historical behavior data and the user behavior path, and obtain service preference information. It can accurately capture the user's needs and preferences, obtain the platform browsing area corresponding to the user behavior path, and push the service preference information to the platform browsing area of ​​the preset target user, which can effectively improve the utilization rate of the user behavior path and the accuracy of recommending personalized content, and finally output the service preference information and feedback it to the user client. The client can be, but is not limited to, various personal computers, laptops, smartphones, tablet computers, and portable wearable devices. The server can be implemented as an independent server or a server cluster consisting of multiple servers. The present invention is described in detail below through specific embodiments.

[0038] The following explains the present invention's description. By performing preference analysis on the target user's historical behavior data and behavior paths, the present invention can accurately capture the user's needs and preferences, thereby providing personalized service recommendations. By performing similarity analysis and convolution operations on the behavior paths, the user's key path features are extracted. The LSTM layer captures temporal dependencies, thereby better understanding the user's dynamic behavior characteristics. Finally, a fully connected layer converts these features into service preference information, improving the utilization of the user's behavior paths and the accuracy of personalized content recommendations.

[0039] Reference Figure 2 FIG. 1 is a flow chart of a method for recommending services based on user behavior according to an embodiment of the present invention. In this embodiment, the method for recommending services based on user behavior includes:

[0040] S1. Obtain the area to be marked on the platform, perform data embedding on the area to be marked, and obtain the embedding area.

[0041] In an embodiment of the present invention, through the overall design of the platform page, different pages are identified, and the pages are further divided to obtain blocks in different pages; for each block, a more detailed pit area is obtained, and the specific content displayed in each pit is obtained, and the pages, blocks, pits and content are summarized as the area to be marked of the platform. Among them, pages: different pages of the platform, such as the homepage, product details page, shopping cart, checkout page, etc.; blocks: different parts within the page, such as the navigation bar, search box, product display area, etc.; pits: more detailed areas within the block, such as advertising space, recommended product space, promotional information box, etc.; content: specific content displayed in each pit, such as product information, advertising content, coupons, etc.

[0042] In a specific healthcare scenario, users navigate from the homepage to the department selection page, then to the doctor list page, and finally to the appointment booking page. Tracking can help track the user's jump path between these different pages. After the user enters the doctor consultation page from the homepage, the system can record the user's behavior in selecting different doctors, submitting questions, or starting a conversation. By tracking the exposure and click data of each consultation step, it is possible to analyze which questions and areas are attracting the most attention.

[0043] In a specific FinTech scenario, a user enters the loan product page from the homepage, selects a different loan product, fills in the application information, and submits it. By tracking every user's actions during the process of selecting a loan product, filling in the application information, and viewing loan details, we can analyze the user's conversion path.

[0044] In an embodiment of the present invention, the step of performing data embedding on the area to be marked to obtain the embedding area includes:

[0045] Obtaining a marking parameter, marking the area to be marked according to the marking parameter, and obtaining a marked area;

[0046] Obtain parameter linking rules, and perform point-linking on the marked area according to the parameter linking rules to obtain a point-linking area.

[0047] In detail, a specific tracking script is embedded in the front-end code of the platform. When the page is loaded, the system will dynamically generate and assign corresponding tracking markup parameters to each page element based on the structure of the page. Tracking markup parameters include: spm, scm, spm_pre, spm_pre2, where spm: the identifier of the page, block, slot, and content, used to identify the specific page, block, or slot where the user is located. scm: the identifier of the goods, usually indicating the identifier of the specific product or advertising content displayed on the page, which is used for subsequent analysis of which products or advertisements are clicked or exposed. spm_pre, spm_pre2: used to record the user's previous and previous-level pages to help track the user's browsing path.

[0048] Combined with the information of the embedding marking parameters, the system can mark each page, block, pit and content with corresponding embedding marking parameters during the development stage. The embedding marking parameters are embedded into the page through code to complete the marking.

[0049] When a user is exposed to or clicks on a certain content, the front-end code will trigger a tracking event and send data containing information such as spm, spm_pre, spm_pre2, and scm to the tracking server. The process from triggering the tracking event to sending the data will be summarized as a parameter link rule.

[0050] Parameters such as spm_pre and spm_pre2 in the page link need to be passed when the page jumps. According to the parameter link rules, the information in the triggered tracking event will be passed to the server, and the data in the platform's unmarked area will be linked to ensure that the user behavior path can be fully tracked and the area where the tracking is completed is obtained.

[0051] By embedding data points in the platform's untagged areas and obtaining embedding tagging parameters, we can accurately capture user behavior and interactions on the platform, thereby forming a user behavioral profile. Furthermore, by embedding and linking the marked areas according to parameter linking rules, we can precisely match user behavior with specific content or functions. This precise data embedding and linking method helps the platform better understand user needs and interests, enabling personalized recommendations, thereby improving recommendation relevance and user experience, and enhancing user stickiness and satisfaction.

[0052] S2. The browsing information of the preset target user in the tracking area is tracked in real time to obtain user behavior data.

[0053] In an embodiment of the present invention, by binding event listeners such as clicks, scrolling, page visits, etc. to the embedding area, the interactive behavior of the target user is tracked and captured in real time to obtain user behavior data.

[0054] In specific medical and health scenarios, by tracking patients' behavior on the hospital's official website or health platform and analyzing which information is frequently accessed (such as departments, doctors, treatment plans, etc.), medical institutions can optimize diagnosis and treatment pathways, reduce the obstacles users encounter during medical treatment, and improve patient satisfaction.

[0055] In specific FinTech scenarios, using tracking technology, financial platforms can monitor users' transactions, logins, capital flows, and other information in real time. When abnormal behavior is detected (such as unusual transaction frequency, large transactions, or changes in geographic location), the system can immediately issue an alert and take necessary anti-fraud measures.

[0056] In an embodiment of the present invention, the real-time tracking of the browsing information of the preset target user in the tracking area to obtain the user behavior data includes:

[0057] Obtain browsing information of preset target users in the tracking area;

[0058] Tracking and capturing the browsing information in real time to obtain a triggering event of the preset target user;

[0059] generating an event type according to the triggering event;

[0060] If the event type is a content click event, then the click element of the preset target user is recorded;

[0061] If the event type is a page access event, the page access sequence and jump history of the preset target user are recorded;

[0062] If the event type is a page scrolling event, recording the page scrolling depth of the preset target user;

[0063] If the event type is a page browsing event, then the time the preset target user stays on the browsing information is recorded;

[0064] The click elements, the page access sequence and jump history, the page scrolling depth and the dwell time are aggregated into user behavior data.

[0065] In detail, when the target user visits the tracking area, the system will record the target user's browsing information based on the tracking mark parameters (such as spm, scm, etc.). The target user's browsing information includes: various different pages of the platform, different elements within the page, etc.

[0066] By using JavaScript to set up event listeners for trigger events, listeners can respond to various types of events, and different event types can be selected to capture user behavior. Event types that trigger events include: clicks on content, page visits, page scrolling, and page views. Assign a callback function to each trigger event type and ensure that the callback function is executed when the event is triggered. The callback function is responsible for processing the event information and collecting and formatting the relevant data before sending it to the server. When an event is triggered, the event objects such as the clicked element, page visit sequence, jump history, page scroll depth, and dwell time are recorded, and the event objects are aggregated into user behavior data.

[0067] By tracking and capturing target users' browsing information in real time, the platform accurately records user behavior data, including click elements, page visit sequence, jump history, scroll depth, and dwell time. This facilitates in-depth analysis of user interests, habits, and needs, providing the platform with a detailed user profile. Based on this behavioral data, the platform can more accurately predict user needs, thereby providing more personalized recommended content and services, improving recommendation accuracy and user satisfaction, ultimately optimizing the user experience and increasing the platform's conversion rate.

[0068] S3. Generate several user behavior paths based on the user behavior data.

[0069] In an embodiment of the present invention, user behavior data is cleaned, and the cleaned data is sorted by time. The sorted user behaviors are classified to construct a number of user behavior paths.

[0070] In healthcare scenarios, by analyzing patient behavior patterns, hospitals can identify frequently accessed symptom and disease queries and provide more targeted medical services based on these behaviors. For example, high-frequency queries for a particular disease can help hospitals determine whether to increase the number of doctors in related departments or launch targeted information promotion campaigns.

[0071] In specific fintech scenarios, users browse different investment products (such as stocks, funds, bonds, etc.) on the platform, check the risk level and return of each product, and may conduct simulated investments. Users may also refer to other users' reviews and even consult investment advisors.

[0072] In an embodiment of the present invention, generating a plurality of user behavior paths according to the user behavior data includes:

[0073] Obtaining a timestamp of the user behavior data, sorting the user behavior data according to the timestamp, and obtaining updated behavior data;

[0074] Extracting the user target behavior of the updated behavior data;

[0075] Construct several user behavior paths based on the user target behavior.

[0076] Specifically, every time a user performs an action (such as clicking, browsing, scrolling, etc.) within a tracking area, the system automatically adds the current time and records a timestamp for the user action data, indicating the specific time when the action occurred. These timestamped user action data are sorted, typically in ascending or descending order based on the timestamp.

[0077] Identify the actions in the update behavior data and use them as the user's target actions. These actions are then linked together according to the specific time they occurred, resulting in a user behavior path within the timestamp. Each user's behavior path consists of multiple consecutive actions, typically represented as a series of events or actions. For example, user A's path might be "browsing the homepage -> clicking a product -> adding it to the shopping cart -> completing the purchase."

[0078] By generating user behavior paths based on user behavior data, the platform can clearly understand the user's interaction process and behavior sequence within a specific time period. Sorting user behavior data by timestamp and extracting target behaviors can help the platform identify changes in user interests and needs. Based on these user behavior paths, the platform can more accurately predict the user's next behavior, further optimize personalized recommendation algorithms, and provide content or services that better meet user preferences, thereby improving the relevance of recommendations and user experience.

[0079] S4. Obtain historical behavior data of a preset target user, perform preference analysis on the preset target user based on the historical behavior data and the user behavior path, and obtain service preference information.

[0080] In this embodiment of the present invention, the platform continuously tracks and records various user interactions to obtain historical behavioral data of target users, including platform log records, clickstream data, user activity tracking, shopping cart records, etc. A convolutional neural network and long-short-term memory network are combined to analyze the target user's historical behavioral data and user behavior paths, predicting the user's desired services and obtaining service preference information.

[0081] In healthcare scenarios, the system analyzes users' health trends based on their physical examination data, data from health monitoring devices (such as smart bracelets), exercise volume, and dietary records. The system can identify health issues such as high blood sugar and high blood pressure, and then push relevant health management suggestions, dietary recommendations, or exercise plans.

[0082] In specific fintech scenarios, the system can recommend the most suitable credit products (such as microloans, credit cards, mortgages, etc.) based on the user's credit score, borrowing history, and repayment ability. For example, if a user has a stable income record and a good credit score, the system may recommend a long-term loan with a low interest rate.

[0083] Figure 3 A flowchart of user preference analysis in a user behavior-based service recommendation method provided by one embodiment of the present invention.

[0084] In an embodiment of the present invention, performing preference analysis on the preset target user based on the historical behavior data and the user behavior path to obtain service preference information includes:

[0085] Performing similarity analysis on a plurality of user behavior paths to obtain path similarity information of the preset target user;

[0086] Convolving the path similarity information to obtain initial path features;

[0087] Performing maximum pooling on the initial path features to obtain key path features;

[0088] Using a preset LSTM layer to capture the time dependency of the key path features and the historical behavior data to obtain user behavior features;

[0089] The user behavior characteristics are fully connected to obtain the service preference information of the preset target user.

[0090] Specifically, in an embodiment of the present invention, performing similarity analysis on a plurality of user behavior paths to obtain path similarity information of the preset target user includes:

[0091] Randomly selecting two user behavior paths from the plurality of user behavior paths, and constructing a path distance matrix using the two selected user behavior paths;

[0092] Obtaining the matrix size of the path distance matrix, and constructing a cumulative cost matrix according to the matrix size;

[0093] Obtaining a filling direction, and filling the cumulative cost matrix according to the filling direction and the path distance matrix to obtain an updated cost matrix;

[0094] Obtaining a backtracking direction, performing a distance search on the updated cost matrix according to the backtracking direction, and obtaining the shortest distance between the two selected user behavior paths;

[0095] Generating similarities between the two selected user behavior paths based on the shortest distance;

[0096] Extracting similar features of the two selected user behavior paths based on the similarity;

[0097] All the similar features are aggregated into the path similarity features of the preset target user.

[0098] In detail, Dynamic Time Warping (DTW) is an algorithm that measures the similarity between two time series. It is particularly suitable for time series with different lengths and different time axis alignments. DTW determines the similarity between two time series by calculating the minimum distance between them.

[0099] Use the two selected user behavior paths to construct a path distance matrix, user behavior paths A and B, where the length of A is m and the length of B is n. Build a distance matrix of size m*n, where each element represents the distance between a data a in A and a data b in B. The calculation formula for each element is as follows:

[0100]

[0101] Among them, a represents the data in user behavior path A, and b represents the data in user behavior path B.

[0102] Construct a cumulative cost matrix of the same size as the distance matrix to record the minimum cumulative distance from the upper left corner of the matrix to the current cell. Initialize the upper left corner element of the matrix to zero: C(0,0) = D(0,0). Get the filling direction, starting from the upper left corner, and fill the cumulative cost matrix according to the filling direction and the path distance matrix, gradually filling the entire matrix. The value of each position is obtained by adding the minimum cumulative distance in the three directions of top, left, and diagonal to the current distance value to obtain the updated cost matrix. The calculation formula is as follows:

[0103] C(i,j)=D(i,j)+min(C(i-1,j),C(i,j-1),C(i-1,j-1))

[0104] Among them, D(i,j) represents the element in the path distance matrix in the i-th row and j-th column, and C(i-1,j), C(i,j-1), and C(i-1,j-1) represent the cumulative distances in the upper, left, and diagonal directions of the cumulative cost matrix.

[0105] Obtain a backtracking direction, starting from the lower-right corner of the matrix. Perform a distance search on the update cost matrix based on the backtracking direction to find a path from the lower-right corner to the upper-left corner. This path corresponds to the optimal alignment between the two time series. The shortest distance between the two selected user behavior paths is calculated as the value of the lower-right element C(m-1,n-1). This shortest distance represents the similarity between the two selected user behavior paths.

[0106] Based on the similarity, similar features are extracted from the two selected user behavior paths, and all of these similar features are aggregated into the target user's path similarity features. For example, a user may frequently browse and inquire about high-value auto insurance products during a certain period, while simultaneously inquiring about and purchasing high-value auto insurance during another period. By comparing these paths, the system discovers that the user has a clear preference for high-value insurance and frequently pays attention to related insurance terms and costs. After aggregating these similar features, the user's "path similarity features" are extracted, indicating that the user tends to choose high-value auto insurance on the car owner insurance platform and frequently consults information on related insurance types. Based on these similar features, the platform can infer the needs of User B and recommend high-value auto insurance products that better meet their needs, or provide personalized insurance discounts and services, thereby improving the user's purchasing experience and the platform's conversion rate.

[0107] By performing cluster analysis on several user behavior paths and extracting similar features of the paths, the platform can accurately identify the behavior patterns and needs of target users, and can efficiently discover potential patterns and similarities from massive amounts of user behavior data, helping the platform to achieve personalized recommendations and precision marketing.

[0108] The function of the convolution layer is to capture local features within different time windows by sliding the convolution kernel. For example, path similarity information is represented by a simple numerical sequence: [1,2,3,4,5,6]. These values ​​represent the similarity of different positions on the path. The preset convolution kernel, for example, a convolution kernel of size 3: [1,0,-1], is used to extract local features of the sequence, such as the trend of change between adjacent points. The convolution kernel and the path similarity information are gradually slid to perform the convolution operation. With a step size of 1 (sliding one position at a time), the final convolution result sequence is: [-2,-2,-2,-2], which obtains the initial path feature.

[0109] Max pooling is a pooling operation commonly used in convolutional neural networks. It aims to extract the most important information from the input feature map. For initial path features, the max pooling operation helps extract key path features. For example, a two-dimensional matrix containing multiple path features represents path information in different regions, such as the time and action of user behavior. A fixed-size sliding window (such as 2x2 or 3x3) is applied to this two-dimensional matrix. The largest value in each window is selected as the representative feature of the window area to obtain the key path feature.

[0110] At the core of LSTM are three gates (input gate, forget gate, and output gate), which work together to determine how the current input features are combined with historical information and how the current state affects the output at the next time step. The LSTM layer enables it to effectively process and memorize dependencies over long time spans. Key path features and historical behavior data are fed into the LSTM layer's input gate, and each piece of data is weighted. The input gate determines how much data to retain in the current memory cell. If a user has clicked on certain items in the past but is no longer interested in them, the forget gate determines whether this outdated information should be discarded to prevent it from affecting current decisions. The output gate uses the current input and internal state (including the results of the input and forget gates) to calculate a new hidden state. This hidden state represents the user's behavioral characteristics at the current moment. The user's current preferences are passed to the fully connected layer through the output gate to infer their service preferences, ultimately obtaining the target user's service preference information.

[0111] By performing preference analysis on target users' historical behavior data and behavior paths, we can accurately capture their needs and preferences, thereby providing personalized service recommendations. By performing similarity analysis and convolution operations on behavior paths, we extract key path features. Using the LSTM layer to capture temporal dependencies, we can better understand users' dynamic behavior characteristics. Finally, through the fully connected layer, these features are converted into service preference information. This improves the utilization of user behavior paths and the accuracy of personalized content recommendations, helping to enhance user experience and increase user satisfaction.

[0112] S5. Obtain the platform browsing area corresponding to the user behavior path, and push the service preference information to the platform browsing area of ​​the preset target user.

[0113] In this embodiment of the present invention, the platform browsing area includes pages without waterfall flow, pages with waterfall flow, video pages, live broadcast pages, etc. The service preference information of the target user obtained through analysis is pushed to the target user, and the user's current platform browsing area is timely inserted with recommendations for preferred products, content, coupons, services, etc., providing users with a one-to-one exclusive intelligent recommendation service.

[0114] In healthcare scenarios, by tracking user behavior on healthcare platforms (such as searching for symptoms, browsing doctor lists, and viewing health reports), we can analyze which medical services, medicines, or health information users are interested in. We can then push medicines or health products suitable for the user's health condition to the area they are browsing, providing convenience for the user.

[0115] In specific FinTech scenarios, platforms can recommend tailored financial products based on user behavior. For example, if a user has browsed stock investment information, the platform can recommend related investment strategies, stock funds, and other products. If a user has visited pages related to auto insurance or mortgages multiple times, the platform can push insurance or loan products that suit their needs to the area they are browsing, improving their user experience.

[0116] In an embodiment of the present invention, obtaining the platform browsing area corresponding to the user behavior path and pushing the service preference information to the platform browsing area of ​​the preset target user includes:

[0117] Obtaining a buried point area in the user behavior path, and generating a platform browsing area of ​​the preset target user based on the buried point area;

[0118] Determine whether there is a preset waterfall flow page in the platform browsing area;

[0119] If the preset waterfall flow page does not exist in the platform browsing area, determining whether the platform browsing area is a video page or a live broadcast page;

[0120] When the platform browsing area is a video page, obtaining the next video page of the video page, and pushing the service preference information to the next video page of the video page;

[0121] When the platform browsing area is a live broadcast page, generating a recommended live broadcast according to the service preference information, and pushing the recommended live broadcast to the next live broadcast page of the live broadcast page;

[0122] When the platform browsing area is not a video page or a live broadcast page, obtaining a drop-down module of the platform browsing area, and adding the service preference information to the drop-down module;

[0123] If the preset waterfall flow page exists in the platform browsing area, determining whether the preset target user is in the waterfall flow block of the preset waterfall flow page;

[0124] When the preset target user is not in the waterfall flow block, obtaining the loading module of the waterfall flow block and adding the service preference information to the loading module;

[0125] When the preset target user is in the waterfall flow block, the service preference information is pushed to the waterfall flow block where the preset target user is.

[0126] Specifically, a waterfall is a flow of information, allowing the page to continuously scroll and load data. The waterfall block is typically at the very bottom of the page and can be pulled up to load. Some pages are pure waterfalls, while others have a waterfall at the bottom.

[0127] The embedding area in the user behavior path, where the embedding mark parameter spm can be used to analyze whether it is a waterfall flow page, and the spm structure is: platform + page + block + pit position.

[0128] ① Video page: Get the next page of the video page, and when cutting the next page, push the service preference information to the next page of the video page, and insert related video recommendations or product / content tile recommendations.

[0129] ② Live broadcast page: Generate recommended live broadcast according to the service preference information, and push the recommended live broadcast to the next live broadcast page of the live broadcast page when the next one is cut.

[0130] ③ No waterfall flow page: The drop-down module in the platform browsing area has the service preference information added to it. For example, if the user wants to learn about accident insurance and the current path is to the Good Car Owner Member Center, then before the next module on the page loads, the user's preferred accident insurance products, accident insurance introduction articles, accident insurance selection videos, and live broadcast recommendations are inserted.

[0131] ④ There is a waterfall page, but it is not in the waterfall block: obtain the loading module of the waterfall block and add the service preference information to the loading module.

[0132] ⑤ There is a waterfall flow page, and in the waterfall flow block: push the service preference information to the waterfall flow block where the target user is, and insert accident insurance products / accident insurance introduction articles / accident insurance selection videos / live broadcast recommendations, etc. analyzed based on user preference needs into the waterfall flow in a timely manner.

[0133] The results of each analysis of the user behavior path are stored. Based on the historical preference results obtained from the analysis of the user behavior path, personalized recommendations for products, content, services, and cards and coupons can be made in the areas of auto insurance, non-auto insurance, services, commodities, content, activities, and cards and coupons, providing users with one-to-one exclusive intelligent recommendation services.

[0134] By accurately identifying the user's browsing area, whether it's a video page, a live stream page, or a waterfall page, we can adopt appropriate push methods based on different page types to ensure that information reaches the page area the user needs. This personalized service push not only improves the user's interactive experience, but also enhances the accuracy of the platform's content recommendations, helping to increase user activity and user stickiness on the platform.

[0135] It should be understood that the size of the serial numbers of the steps in the above embodiments 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 invention.

[0136] like Figure 4 FIG. 1 is a functional module diagram of a service recommendation device based on user behavior provided by an embodiment of the present invention.

[0137] In an embodiment of the present disclosure, a service recommendation device based on user behavior is provided, and the service recommendation device based on user behavior corresponds to the service recommendation method based on user behavior in the above embodiment. Figure 4 As shown, the service recommendation device 100 based on user behavior can be installed in an electronic device. According to the functions to be implemented, the service recommendation device 100 based on user behavior includes a data tracking module 101, a data tracking module 102, a path generation module 103, a preference analysis module 104, and a service recommendation module 105. The functional modules are described in detail as follows:

[0138] The data embedding module 101 is used to obtain the area to be marked on the platform, and to embed data in the area to be marked to obtain the embedding area;

[0139] The data tracking module 102 is used to track the browsing information of the preset target user in the tracking area in real time to obtain user behavior data;

[0140] A path generation module 103 is configured to generate a plurality of user behavior paths based on the user behavior data;

[0141] The preference analysis module 104 is used to obtain historical behavior data of a preset target user, perform preference analysis on the preset target user based on the historical behavior data and the user behavior path, and obtain service preference information;

[0142] The service recommendation module 105 is configured to obtain the platform browsing area corresponding to the user behavior path and push the service preference information to the platform browsing area of ​​the preset target user.

[0143] In one embodiment, when the data embedding module 101 acquires the area to be marked of the platform and embeds data in the area to be marked to obtain the embedding area, it is used to:

[0144] Obtaining a marking parameter, marking the area to be marked according to the marking parameter, and obtaining a marked area;

[0145] Obtain parameter linking rules, and perform point-linking on the marked area according to the parameter linking rules to obtain a point-linking area.

[0146] In one embodiment, when the data tracking module 102 performs real-time tracking of browsing information of a preset target user in the tracking area to obtain user behavior data, it is used to:

[0147] Obtain browsing information of preset target users in the tracking area;

[0148] Tracking and capturing the browsing information in real time to obtain a triggering event of the preset target user;

[0149] generating an event type according to the triggering event;

[0150] If the event type is a content click event, then the click element of the preset target user is recorded;

[0151] If the event type is a page access event, the page access sequence and jump history of the preset target user are recorded;

[0152] If the event type is a page scrolling event, recording the page scrolling depth of the preset target user;

[0153] If the event type is a page browsing event, then the time the preset target user stays on the browsing information is recorded;

[0154] The click elements, the page access sequence and jump history, the page scrolling depth and the dwell time are aggregated into user behavior data.

[0155] In one embodiment, when generating a plurality of user behavior paths based on the user behavior data, the path generation module 103 is configured to:

[0156] Obtaining a timestamp of the user behavior data, sorting the user behavior data according to the timestamp, and obtaining updated behavior data;

[0157] Extracting the user target behavior of the updated behavior data;

[0158] Construct several user behavior paths based on the user target behavior.

[0159] In one embodiment, when the preference analysis module 104 performs a preference analysis on the preset target user based on the historical behavior data and the user behavior path to obtain service preference information, it is configured to:

[0160] Performing similarity analysis on a plurality of user behavior paths to obtain path similarity information of the preset target user;

[0161] Convolving the path similarity information to obtain initial path features;

[0162] Performing maximum pooling on the initial path features to obtain key path features;

[0163] Using a preset LSTM layer to capture the time dependency of the key path features and the historical behavior data to obtain user behavior features;

[0164] The user behavior characteristics are fully connected to obtain the service preference information of the preset target user.

[0165] In one embodiment, when the preference analysis module 104 performs a preference analysis on the preset target user based on the historical behavior data and the user behavior path to obtain service preference information, it is configured to:

[0166] Randomly selecting two user behavior paths from the plurality of user behavior paths, and constructing a path distance matrix using the two selected user behavior paths;

[0167] Obtaining the matrix size of the path distance matrix, and constructing a cumulative cost matrix according to the matrix size;

[0168] Obtaining a filling direction, and filling the cumulative cost matrix according to the filling direction and the path distance matrix to obtain an updated cost matrix;

[0169] Obtaining a backtracking direction, performing a distance search on the updated cost matrix according to the backtracking direction, and obtaining the shortest distance between the two selected user behavior paths;

[0170] Generating similarities between the two selected user behavior paths based on the shortest distance;

[0171] Extracting similar features of the two selected user behavior paths based on the similarity;

[0172] All the similar features are aggregated into the path similarity features of the preset target user.

[0173] In one embodiment, when the service recommendation module 105 obtains the platform browsing area corresponding to the user behavior path and pushes the service preference information to the platform browsing area of ​​the preset target user, it is used to:

[0174] Obtaining a buried point area in the user behavior path, and generating a platform browsing area of ​​the preset target user based on the buried point area;

[0175] Determine whether there is a preset waterfall flow page in the platform browsing area;

[0176] If the preset waterfall flow page does not exist in the platform browsing area, determining whether the platform browsing area is a video page or a live broadcast page;

[0177] When the platform browsing area is a video page, obtaining the next video page of the video page, and pushing the service preference information to the next video page of the video page;

[0178] When the platform browsing area is a live broadcast page, generating a recommended live broadcast according to the service preference information, and pushing the recommended live broadcast to the next live broadcast page of the live broadcast page;

[0179] When the platform browsing area is not a video page or a live broadcast page, obtaining a drop-down module of the platform browsing area, and adding the service preference information to the drop-down module;

[0180] If the preset waterfall flow page exists in the platform browsing area, determining whether the preset target user is in the waterfall flow block of the preset waterfall flow page;

[0181] When the preset target user is not in the waterfall flow block, obtaining the loading module of the waterfall flow block and adding the service preference information to the loading module;

[0182] When the preset target user is in the waterfall flow block, the service preference information is pushed to the waterfall flow block where the preset target user is.

[0183] In the present invention, for a service recommendation device based on user behavior, first, the present invention obtains the area to be marked of the platform, performs data embedding on the area to be marked, and obtains the embedding area. By embedding data on the area to be marked of the platform and obtaining embedding marking parameters, the user's behavior and interaction on the platform can be accurately captured, and the browsing information of the preset target user in the embedding area is tracked in real time to obtain user behavior data. Based on these behavior data, the platform can more accurately predict the user's needs, generate several user behavior paths according to the user behavior data, and clearly understand the user's interaction process and behavior sequence in a specific time period, obtain the historical behavior data of the preset target user, and then perform preference analysis on the preset target user according to the historical behavior data and the user behavior path to obtain service preference information, accurately capture the user's needs and preferences, obtain the platform browsing area corresponding to the user behavior path, and finally push the service preference information to the platform browsing area of ​​the preset target user, which can effectively improve the utilization rate of the user behavior path and the accuracy of recommending personalized content. For the specific definition of a service recommendation device based on user behavior, please refer to the definition of a service recommendation method based on user behavior above, which will not be repeated here. Each module in the aforementioned user behavior-based service recommendation device may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor in a computer device in the form of hardware, or may be stored in a memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0184] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, it realizes the functions or steps on the service side of a service recommendation method based on user behavior.

[0185] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as follows: Figure 6As shown. The computer device includes a processor, memory, network interface, display screen and input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements the functions or steps on the client side of a service recommendation method based on user behavior.

[0186] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the following steps are performed:

[0187] Obtaining the area to be marked on the platform, performing data embedding on the area to be marked, and obtaining the embedding area;

[0188] Real-time tracking of the browsing information of the preset target user in the tracking area to obtain user behavior data;

[0189] generating a plurality of user behavior paths according to the user behavior data;

[0190] Obtaining historical behavior data of a preset target user, performing preference analysis on the preset target user based on the historical behavior data and the user behavior path, and obtaining service preference information;

[0191] The platform browsing area corresponding to the user behavior path is obtained, and the service preference information is pushed to the platform browsing area of ​​the preset target user.

[0192] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and apparatuses can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the module division is merely a logical function division, and actual implementation may employ other division methods.

[0193] In addition, the functional modules in various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or hardware plus software functional modules.

[0194] Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the claims are intended to be embraced therein. Any reference to a figure in a claim should not be construed as limiting the claim to which it relates.

[0195] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0196] In some implementations of this embodiment, a computer-readable storage medium is provided, on which a computer program is stored, characterized in that when the computer program is executed by a processor, the steps of the method described in the above embodiment are implemented.

[0197] The readable storage medium of the present invention stores a computer program, which, when executed by a processor of an electronic device, can implement:

[0198] Obtaining the area to be marked on the platform, performing data embedding on the area to be marked, and obtaining the embedding area;

[0199] Real-time tracking of the browsing information of the preset target user in the tracking area to obtain user behavior data;

[0200] generating a plurality of user behavior paths according to the user behavior data;

[0201] Obtaining historical behavior data of a preset target user, performing preference analysis on the preset target user based on the historical behavior data and the user behavior path, and obtaining service preference information;

[0202] The platform browsing area corresponding to the user behavior path is obtained, and the service preference information is pushed to the platform browsing area of ​​the preset target user.

[0203] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can be found in the relevant descriptions of the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0204] The computer-readable storage medium may also store at least one computer-executable program / instruction, such as a computer-readable instruction. Computer-readable storage media include, but are not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Computer-readable storage media may include, for example, read-only memory (ROM), a hard disk, a flash memory, etc. For example, a non-transitory computer-readable storage medium may be connected to a computing device such as a computer, and then, when the computing device executes the computer-readable instructions stored on the computer-readable storage medium, the various methods described above may be performed.

[0205] In addition, the computer device may also include (but is not limited to) a data bus, an input / output (I / O) bus, a display, and input / output devices (eg, keyboard, mouse, speaker, etc.).

[0206] The processor can communicate with external devices via an I / O bus via a wired or wireless network.

[0207] In one embodiment, the at least one computer executable instruction may also be compiled into or constitute a software product / computer program product, wherein one or more computer executable instructions are executed by a processor to perform the various functions and / or method steps in the embodiments described in the present technology.

[0208] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0209] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0210] In the embodiments provided in the present disclosure, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a portion of code, and the above-mentioned module, program segment or a portion of code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or can be implemented using a combination of dedicated hardware and computer instructions.

[0211] It should be noted that, in this disclosure, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element limited by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0212] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

[0213] It should be noted that if software tools or components other than those of our company appear in the embodiments of this application, they are only used for illustration and do not represent actual use.

Claims

1. A service recommendation method based on user behavior, characterized in that: The method comprises: Obtaining the area to be marked on the platform, performing data embedding on the area to be marked, and obtaining the embedding area; Track the browsing information of preset target users in the tracking area in real time to obtain user behavior data; generating a plurality of user behavior paths according to the user behavior data; Obtaining historical behavior data of a preset target user, performing preference analysis on the preset target user based on the historical behavior data and the user behavior path, and obtaining service preference information; The platform browsing area corresponding to the user behavior path is obtained, and the service preference information is pushed to the platform browsing area of ​​the preset target user.

2. The service recommendation method based on user behavior according to claim 1, characterized in that: The step of performing data embedding on the area to be marked to obtain the embedding area includes: Obtaining a marking parameter, marking the area to be marked according to the marking parameter, and obtaining a marked area; Obtain parameter linking rules, and perform point-linking on the marked area according to the parameter linking rules to obtain a point-linking area.

3. The service recommendation method based on user behavior according to claim 1, characterized in that: The real-time tracking of the browsing information of the preset target user in the tracking area to obtain user behavior data includes: Obtain browsing information of preset target users in the tracking area; Tracking and capturing the browsing information in real time to obtain a triggering event of the preset target user; generating an event type according to the triggering event; If the event type is a content click event, then the click element of the preset target user is recorded; If the event type is a page access event, the page access sequence and jump history of the preset target user are recorded; If the event type is a page scrolling event, recording the page scrolling depth of the preset target user; If the event type is a page browsing event, then the time the preset target user stays on the browsing information is recorded; The click elements, the page access sequence and jump history, the page scrolling depth and the dwell time are aggregated into user behavior data.

4. The service recommendation method based on user behavior according to claim 1, characterized in that: Generating a plurality of user behavior paths according to the user behavior data includes: Obtaining a timestamp of the user behavior data, sorting the user behavior data according to the timestamp, and obtaining updated behavior data; Extracting the user target behavior of the updated behavior data; Construct several user behavior paths based on the user target behavior.

5. The service recommendation method based on user behavior according to claim 1, characterized in that: The performing preference analysis on the preset target user based on the historical behavior data and the user behavior path to obtain service preference information includes: Performing similarity analysis on a plurality of user behavior paths to obtain path similarity information of the preset target user; Convolving the path similarity information to obtain initial path features; Performing maximum pooling on the initial path features to obtain key path features; Using a preset LSTM layer to capture the time dependency of the key path features and the historical behavior data to obtain user behavior features; The user behavior characteristics are fully connected to obtain the service preference information of the preset target user.

6. The service recommendation method based on user behavior according to claim 5, characterized in that: The performing similarity analysis on the plurality of user behavior paths to obtain path similarity information of the preset target user includes: Randomly selecting two user behavior paths from the plurality of user behavior paths, and constructing a path distance matrix using the two selected user behavior paths; Obtaining the matrix size of the path distance matrix, and constructing a cumulative cost matrix according to the matrix size; Obtaining a filling direction, and filling the cumulative cost matrix according to the filling direction and the path distance matrix to obtain an updated cost matrix; Obtaining a backtracking direction, performing a distance search on the updated cost matrix according to the backtracking direction, and obtaining the shortest distance between the two selected user behavior paths; Generating similarities between the two selected user behavior paths based on the shortest distance; Extracting similar features of the two selected user behavior paths based on the similarity; All the similar features are aggregated into the path similarity features of the preset target user.

7. The service recommendation method based on user behavior according to claim 1, characterized in that: The acquiring the platform browsing area corresponding to the user behavior path and pushing the service preference information to the platform browsing area of ​​the preset target user includes: Obtaining a buried point area in the user behavior path, and generating a platform browsing area of ​​the preset target user based on the buried point area; Determine whether there is a preset waterfall flow page in the platform browsing area; If the preset waterfall flow page does not exist in the platform browsing area, determining whether the platform browsing area is a video page or a live broadcast page; When the platform browsing area is a video page, obtaining the next video page of the video page, and pushing the service preference information to the next video page of the video page; When the platform browsing area is a live broadcast page, generating a recommended live broadcast according to the service preference information, and pushing the recommended live broadcast to the next live broadcast page of the live broadcast page; When the platform browsing area is not a video page or a live broadcast page, obtaining a drop-down module of the platform browsing area, and adding the service preference information to the drop-down module; If the preset waterfall flow page exists in the platform browsing area, determining whether the preset target user is in the waterfall flow block of the preset waterfall flow page; When the preset target user is not in the waterfall flow block, obtaining the loading module of the waterfall flow block and adding the service preference information to the loading module; When the preset target user is in the waterfall flow block, the service preference information is pushed to the waterfall flow block where the preset target user is.

8. A service recommendation device based on user behavior, characterized in that: The device comprises: The data embedding module is used to obtain the area to be marked on the platform, embed data in the area to be marked, and obtain the embedding area; A data tracking module is used to track the browsing information of preset target users in the tracking area in real time to obtain user behavior data; A path generation module, configured to generate a plurality of user behavior paths based on the user behavior data; A preference analysis module is used to obtain historical behavior data of a preset target user, perform preference analysis on the preset target user based on the historical behavior data and the user behavior path, and obtain service preference information; The service recommendation module is used to obtain the platform browsing area corresponding to the user behavior path and push the service preference information to the platform browsing area of ​​the preset target user.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the service recommendation method based on user behavior as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for recommending services based on user behavior as claimed in any one of claims 1 to 7 is implemented.

Images