User recommendation method and device, electronic equipment and storage medium
By screening candidate users in social applications based on user profile and security level models, and recommending them in combination with social probability, the problem of risky user recommendations in the existing technology is solved, and more accurate user recommendations are achieved.
Patent Information
- Application Number
- CN202510191408.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-20
- Publication Date
- 2025-07-11
AI Technical Summary
When existing social applications recommend users, they are prone to recommending risk users, resulting in user interference and lack effective screening and recommendation mechanisms.
Based on the user portrait of the target user, candidate users are filtered from multiple registered users, the security level value of candidate users is evaluated through the security level model, and social probability is predicted based on historical social behavior, and recommendations are made only when a certain security level and social probability are met.
It effectively avoids the risk of recommending users to target users, improves the pertinence and accuracy of user recommendations, and reduces unnecessary interference.
Smart Images
Figure CN120296053A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of network social technologies, and in particular, to a user recommendation method, apparatus, electronic device, and storage medium. Background Art
[0002] In common social applications (APPs), after a user completes registration, the service system often recommends some other users to the user, so as to prompt the user to send a friend application to the users of interest, and thus promote adding friends for interaction.
[0003] Then, when the system recommends other users to the user, if there are risky users among the recommended users, it will cause unnecessary interference to the user. Therefore, how to better perform user recommendation is an urgent problem for those skilled in the art. Summary of the Invention
[0004] This application provides a user recommendation method, apparatus, electronic device, and storage medium, which are used to solve the problem of how to better perform user recommendation in the prior art.
[0005] In a first aspect, this application provides a user recommendation method, including:
[0006] When the target user is a safe user, based on the user profile of the target user, screening multiple candidate users from multiple registered users;
[0007] For each candidate user, based on the registration information and behavior information of the candidate user, determining the security level value of the candidate user; when the security level value is greater than a preset level value, predicting the social probability of the candidate user and the target user for social interaction based on the historical social behavior of the candidate user;
[0008] When the social probability corresponding to the candidate user is greater than a preset probability, recommending the candidate user to the target user.
[0009] According to a user recommendation method provided by this application, the user profile includes location information, interest preferences, social circles, and behavior information. The screening of multiple candidate users from multiple registered users based on the user profile of the target user includes:
[0010] Respectively determining the weights corresponding to the location information, interest preferences, social circles, and behavior information of the target user; where the weights are used to represent the attention degree of the target user to the location information, interest preferences, social circles, and behavior information respectively;
[0011] For each registered user, respectively determine the first matching degree between the location information of the target user and the location information of the registered user, the second matching degree between the interest preferences of the target user and the interest preferences of the registered user, the third matching degree between the social circle of the target user and the social circle of the registered user, and the fourth matching degree between the behavior information of the target user and the behavior information of the registered user; and based on the first matching degree, the second matching degree, the third matching degree, the fourth matching degree, and the weights corresponding to the location information, interest preferences, social circle, and behavior information respectively, determine the target matching degree between the target user and the registered user;
[0012] Based on the target matching degree between the target user and each registered user, screen the multiple candidate users from the multiple registered users.
[0013] According to a user recommendation method provided by the present application, the method further includes:
[0014] Obtain the location information, interest preferences, social circle, and behavior information of the first candidate user who is a friend added by the target user among all the recommended candidate users;
[0015] Match the location information, interest preferences, social circle, and behavior information of the target user with the location information, interest preferences, social circle, and behavior information of the first candidate user respectively to obtain a first matching result;
[0016] Based on the first matching result, adjust the weights corresponding to the location information, interest preferences, social circle, and behavior information of the target user respectively.
[0017] According to a user recommendation method provided by the present application, the adjusting the weights corresponding to the location information, interest preferences, social circle, and behavior information of the target user respectively based on the first matching result includes:
[0018] When at least one target information in the location information, interest preferences, social circle, and behavior information of the target user is not successfully matched in the first matching result, match the target information of the target user with the target information of the second candidate user to obtain a second matching result; wherein, the second candidate user is a candidate user who is not a friend added by the target user among all the recommended candidate users;
[0019] When the second matching result indicates successful matching, reduce the weights of each target information, and increase the weights of the other information except the target information in the location information, interest preferences, social circle, and behavior information of the target user.
[0020] A user recommendation method provided by the present application, the step of recommending the candidate user to the target user when the social probability corresponding to the candidate user is greater than a preset probability includes:
[0021] When the social probability corresponding to the candidate user is greater than a preset probability, obtain the historical dissemination content of the candidate user;
[0022] Perform sentiment analysis on the candidate user based on the historical dissemination content to obtain a sentiment analysis result; wherein, the sentiment analysis result includes positive sentiment, neutral sentiment or negative sentiment;
[0023] When the sentiment analysis result is positive sentiment, recommend the candidate user to the target user.
[0024] A user recommendation method provided by the present application, the method further includes:
[0025] Monitor the behavior information of the first candidate user who has been added as a friend by the target user;
[0026] Based on the registration information and behavior information of the first candidate user, re-determine the security level value of the first candidate user;
[0027] When the security level value of the first candidate user is less than or equal to the preset level value, output a prompt message to the terminal corresponding to the target user, and the prompt message is used to remind that the first candidate user is a risky user.
[0028] A user recommendation method provided by the present application, the step of determining the security level value of the candidate user based on the registration information and behavior information of the candidate user includes:
[0029] Input the registration information and behavior information of the candidate user into a security level model to obtain the security level value output by the security level model;
[0030] Wherein, the security level model is obtained by training an initial security level model based on the sample registration information, sample behavior information of multiple sample users, and the security level value labels corresponding to the multiple sample users respectively.
[0031] In a second aspect, the present application further provides a user recommendation device, including:
[0032] A screening unit, configured to screen multiple candidate users from multiple registered users based on the user profile of the target user when the target user is a secure user;
[0033] A first processing unit is configured to determine a security level value for each candidate user based on the registration information and behavior information of the candidate user; and in the case where the security level value is greater than a preset level value, predict a social probability that the candidate user socializes with the target user based on the historical social behavior of the candidate user.
[0034] A recommendation unit is configured to recommend the candidate user to the target user in the case where the social probability corresponding to the candidate user is greater than a preset probability.
[0035] The present application further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where when the processor executes the computer program, the user recommendation method described in any one of the above is implemented.
[0036] The present application further provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the user recommendation method described in any one of the above is implemented.
[0037] The present application further provides a computer program product, including a computer program, and when the computer program is executed by a processor, the user recommendation method described in any one of the above is implemented.
[0038] The user recommendation method, device, electronic device, and storage medium provided by the present application, in the case where the target user is a secure user, screen multiple candidate users from multiple registered users based on the user profile of the target user; for each candidate user, determine the security level value of the candidate user based on the registration information and behavior information of the candidate user; in the case where the security level value is greater than a preset level value, predict the social probability that the candidate user socializes with the target user based on the historical social behavior of the candidate user; and recommend the candidate user to the target user in the case where the social probability corresponding to the candidate user is greater than a preset probability. In this way, by combining the security level value of the candidate user and the corresponding social probability, candidate user recommendation is jointly performed, which can not only effectively avoid recommending risky users to the target user, but also perform user recommendation more pertinently. Description of the Drawings
[0039] In order to more clearly illustrate the technical solutions in the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.
[0040] Figure 1 It is a schematic flowchart of a user recommendation method provided by an embodiment of the present application.
[0041] Figure 2 This is a schematic flowchart of a process for screening multiple candidate users from multiple registered users provided by an embodiment of the present application.
[0042] Figure 3 This is a schematic structural diagram of a user recommendation device provided by an embodiment of the present application.
[0043] Figure 4 This is a schematic physical structure diagram of an electronic device provided by an embodiment of the present application. Detailed implementation manners
[0044] To make the objectives, technical solutions, and advantages of the present application clearer, the technical solutions in the present application will be clearly and completely described below with reference to the accompanying drawings in the present application. Apparently, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without making creative efforts shall fall within the protection scope of the present application.
[0045] In the embodiments of the present application, "at least one" means one or more, and "multiple" means two or more. "And / or" describes the association relationship of associated objects and indicates that three relationships may exist. For example, A and / or B may represent three situations: A exists alone, A and B exist simultaneously, and B exists alone, where A and B may be singular or plural. In the text description of the present application, the character " / " generally represents an "or" relationship between the associated objects before and after.
[0046] The technical solutions provided by the embodiments of the present application can be applied to the user recommendation scenario. In a general social APP, after a user completes registration, the service system often recommends some other users to the user, so as to prompt the user to send a friend application to the users of interest, thereby promoting adding friends for interaction.
[0047] Then, when the service system recommends other users to the user, if there are risk users among the recommended users, it will cause unnecessary interference to the user.
[0048] To perform user recommendation better, the embodiments of the present application provide a user recommendation method. Next, the user recommendation method provided by the present application will be described in detail through the following several specific embodiments. It can be understood that the following several specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.
[0049] Figure 1 This is a schematic flowchart of a user recommendation method provided by an embodiment of the present application. For example, please refer toFigure 1 As shown, the user recommendation method may include:
[0050] S101. When the target user is a secure user, based on the user profile of the target user, screen multiple candidate users from multiple registered users.
[0051] Exemplarily, in the embodiments of the present application, the user profile may include location information, interest preferences, social circles, behavior information, etc., which can be specifically set according to actual needs.
[0052] It can be understood that in the embodiments of the present application, when making user recommendations for the target user, it is first determined that the target user is a secure user before making user recommendations for it, which can effectively avoid recommending other secure users to risk users, thereby effectively reducing the risk of other secure users being recommended.
[0053] Exemplarily, when determining whether the target user is a secure user, it can be determined whether the target user is a secure user based on the registration information and behavior information of the target user. When it is determined that the target user is a secure user, then based on the user profile of the target user, screen multiple candidate users from multiple registered users, that is, to achieve the preliminary screening of user recommendations.
[0054] Exemplarily, the registration information may include the user's basic identity information, verification information, or social management information, etc. Among them, the basic identity information, such as name, gender, contact information, etc.; the verification information, such as email verification, mobile phone verification, real-name authentication status, etc.; the social management information, such as whether other social media accounts are associated, etc.
[0055] Exemplarily, the behavior information may be the user's login behavior, operation behavior, or social interaction behavior, etc. Among them, the login behavior, such as login frequency, login location, login device type, etc.; the operation behavior mainly refers to the activity track on the application program, such as browsing, clicking, searching, purchasing, commenting, etc.; the social interaction mainly refers to the communication situation with other users, such as private messages, comment replies, etc.
[0056] After screening out multiple candidate users from multiple registered users, the following S102 and S103 can be executed to achieve the re-screening of user recommendations.
[0057] S102. For each candidate user, based on the registration information and behavior information of the candidate user, determine the security level value of the candidate user; when the security level value is greater than the preset level value, based on the historical social behavior of the candidate user, predict the social probability of the candidate user interacting with the target user.
[0058] Among them, the value of the preset level can be set according to actual needs, and specific limitations are not imposed in the embodiments of the present application here.
[0059] Exemplarily, in the embodiments of the present application, when determining the security level value of a candidate user based on the registration information and behavior information of the candidate user, a security level model based on deep learning can be used to input the registration information and behavior information of the candidate user into the security level model, and the security level value output by the security level model can be obtained.
[0060] Among them, the security level model is obtained by training an initial security level model based on the sample registration information, sample behavior information of multiple sample users, and the security level value labels corresponding to the multiple sample users respectively.
[0061] Exemplarily, the security level model can be a Convolutional Neural Network (CNN) model, or a Recurrent Neural Network (RNN) model, or a Self-Attention Mechanism model, etc., and can be specifically set according to actual needs.
[0062] It can be understood that when determining the security level value of a candidate user based on the registration information and behavior information of the candidate user, in addition to using a security level model based on deep learning, other methods can also be adopted. For example, a security level algorithm can be used to calculate the security level value of the candidate user, which can be specifically set according to actual needs.
[0063] In the case where the security level value of the candidate user is greater than the preset level value, it indicates that the candidate user is a secure user. In this way, by calculating the security level value of the candidate user, the risk of recommending a risky user to the target user can be effectively avoided. In addition, in the embodiments of the present application, in order to perform user recommendation more pertinently, the social probability of the candidate user interacting with the target user can be further predicted based on the historical social behavior of the candidate user, and when the social probability corresponding to the candidate user is greater than the preset probability, the candidate user can be recommended to the target user, that is, the following S103 is executed, so that user recommendation can be more targeted.
[0064] Exemplarily, in the embodiments of the present application, the historical social behavior of a user can include the user's attention volume, interaction frequency, interaction content, or the number of mutual friends, the tightness of the social relationship, etc., which can be specifically set according to actual needs.
[0065] Exemplarily, in the embodiments of the present application, when predicting the social probability of a candidate user interacting with a target user based on the historical social behavior of the candidate user, the historical social behavior of the candidate user and the historical social behavior of the target user can be input into a pre-trained social probability prediction model, and the social probability prediction model is used to predict the social probability of the candidate user interacting with the target user, so as to determine the social probability corresponding to the candidate user.
[0066] S103. When the social probability corresponding to the candidate user is greater than a preset probability, recommend the candidate user to the target user.
[0067] Among them, the value of the preset probability can be set according to actual needs. Here, the specific value of the preset probability is not further limited in the embodiments of the present application.
[0068] It can be seen that in the embodiments of the present application, when making user recommendations for the target user, when the target user is a safe user, based on the user profile of the target user, multiple candidate users can be screened from multiple registered users; for each candidate user, based on the registration information and behavior information of the candidate user, determine the security level value of the candidate user; when the security level value is greater than a preset level value, based on the historical social behavior of the candidate user, predict the social probability of the candidate user interacting with the target user; and when the social probability corresponding to the candidate user is greater than a preset probability, recommend the candidate user to the target user. In this way, by combining the security level value of the candidate user and its corresponding social probability to jointly recommend candidate users, not only can the risk of recommending risk users to the target user be effectively avoided, but also user recommendations can be made more pertinently.
[0069] Based on the above Figure 1 shown embodiments, taking the user profile including location information, interest preferences, social circles, and behavior information as an example, exemplarily, in the above S101, when screening multiple candidate users from multiple registered users based on the user profile of the target user, reference can be made to Figure 2 shown, Figure 2 is a schematic flow chart of a method for screening multiple candidate users from multiple registered users provided by an embodiment of the present application. The method may include:
[0070] S201. Respectively determine the weights corresponding to the location information, interest preferences, social circles, and behavior information of the target user.
[0071] Among them, the weight is used to represent the degree of attention of the target user to the location information, interest preferences, social circles, and behavior information respectively.
[0072] Specifically, the weight corresponding to the location information is used to represent the degree of attention of the target user to the location information, the weight corresponding to the interest preference is used to represent the degree of attention of the target user to the interest preference, the weight corresponding to the social circle is used to represent the degree of attention of the target user to the social circle, and the weight corresponding to the behavior information is used to represent the degree of attention of the target user to the behavior information.
[0073] Exemplarily, when respectively determining the weights corresponding to the location information, interest preference, social circle and behavior information of the target user, if the target user often added some users in the same location before, the weight of the location information can be set higher; if the target user often added some users with the same interest preference before, the weight of the interest preference can be set higher; if the target user often added some users in the same social circle before, the weight of the social circle can be set higher; if the target user often added some users with similar behavior information before, the weight of the behavior information can be set higher, and it can be specifically set according to actual needs.
[0074] S202. For each registered user, respectively determine the first matching degree between the location information of the target user and the location information of the registered user, the second matching degree between the interest preference of the target user and the interest preference of the registered user, the third matching degree between the social circle of the target user and the social circle of the registered user, and the fourth matching degree between the behavior information of the target user and the behavior information of the registered user; and based on the first matching degree, the second matching degree, the third matching degree, the fourth matching degree, and the weights corresponding to the location information, interest preference, social circle and behavior information respectively, determine the target matching degree between the target user and the registered user.
[0075] Exemplarily, when determining the target matching degree between the target user and the registered user based on the first matching degree, the second matching degree, the third matching degree, the fourth matching degree, and the weights corresponding to the location information, interest preference, social circle and behavior information respectively, the product of the first matching degree and the weight corresponding to the location information can be calculated, the product of the second matching degree and the weight corresponding to the interest preference can be calculated, the product of the third matching degree and the weights corresponding to the social circle respectively can be calculated, and the product of the fourth matching degree and the weights corresponding to the behavior information respectively can be calculated, and the sum value of these four products is determined as the target matching degree between the target user and the registered user.
[0076] After determining the target matching degree between the target user and each registered user, the following S203 can be executed:
[0077] S203. Based on the target matching degree between the target user and each registered user, screen multiple candidate users from multiple registered users.
[0078] For example, when screening multiple candidate users from multiple registered users based on the target matching degree between the target user and each registered user, the top M registered users corresponding to the target matching degrees in descending order of the target matching degree can be determined as the multiple candidate users; alternatively, the registered users corresponding to the target matching degrees greater than a certain matching degree threshold can be determined as the multiple candidate users, etc., which can be specifically set according to actual needs.
[0079] By determining the weights corresponding to the location information, interest preferences, social circles, and behavior information of the target user respectively, and based on the first matching degree, the second matching degree, the third matching degree, the fourth matching degree, and the weights corresponding to the location information, interest preferences, social circles, and behavior information respectively, to determine the target matching degree between the target user and the registered users, and then screening multiple candidate users from multiple registered users based on the target matching degree between the target user and each registered user, the preliminary screening of the recommended users can be achieved.
[0080] It should be noted that in the embodiment of the present application, in the above S201, the weights corresponding to the location information, interest preferences, social circles, and behavior information of the target user are dynamically set and can be set according to actual needs.
[0081] For example, in the embodiment of the present application, after recommending candidate users to the target user, the location information, interest preferences, social circles, and behavior information of the first candidate user who has been added as a friend by the target user among all the recommended candidate users can also be obtained; and the location information, interest preferences, social circles, and behavior information of the target user are respectively matched with the location information, interest preferences, social circles, and behavior information of the first candidate user to obtain a first matching result; based on the first matching result, the weights corresponding to the location information, interest preferences, social circles, and behavior information of the target user are adjusted, realizing the dynamic adjustment of the weights, thereby improving the accuracy of user recommendation.
[0082] For example, assume that the target user adds the first candidate user A as a friend. If the first matching result indicates that both the location information and interest preferences of the first candidate user A are the same as those of the target user, it means that the target user pays a higher attention to the location information and interest preferences. In this case, the weights of the location information and interest preferences can be increased; if the first matching result indicates that both the location information and social circle of the first candidate user A are the same as those of the target user, it means that the target user pays a higher attention to the location information and social circle. In this case, the weights of the location information and social circle can be increased. In this way, through the first matching result, the weights of the matching user information can be appropriately increased.
[0083] Exemplarily, when adjusting the weights corresponding to the location information, interest preferences, social circles, and behavior information of the target user based on the first matching result, in the case where the first matching result indicates that at least one target information among the location information, interest preferences, social circles, and behavior information of the target user fails to match successfully, the target information of the target user is matched with the target information of the second candidate user to obtain a second matching result; wherein, the second candidate user is a candidate user among all the recommended candidate users who has not been added as a friend by the target user; in the case where the second matching result indicates a successful match, the weights of each target information are reduced, and the weights of the other information except the target information among the location information, interest preferences, social circles, and behavior information of the target user are increased.
[0084] Exemplarily, assume that the target user has added the first candidate user A as a friend. If the first matching result indicates that the location information and interest preferences of the first candidate user A are the same as those of the target user, and the social circle of the first candidate user A is different from that of the target user, that is, the social circle of the first candidate user A does not match the social circle of the target user. In this case, it is impossible to determine the degree of attention of the target user to the social circle. Therefore, the social circle of the target user can be further matched with the social circle of the second candidate user B. If the second matching result indicates that the social circle of the target user is the same as the social circle of the second candidate user B who has not been added as a friend, it means that the target user has a relatively low degree of attention to the social circle. In this case, the weight of the social circle can be reduced, and the weights of the location information and interest preferences can be increased, realizing the dynamic adjustment of the weights, thereby improving the accuracy of user recommendation.
[0085] Based on any of the above embodiments, exemplarily, in the above S103, when recommending a candidate user to the target user in the case where the social probability corresponding to the candidate user is greater than the preset probability, the historical dissemination content of the candidate user can also be obtained first; and based on the historical dissemination content, a sentiment analysis is performed on the candidate user to obtain a sentiment analysis result; wherein, the sentiment analysis result includes positive sentiment, neutral sentiment, or negative sentiment; in the case where the sentiment analysis result is positive sentiment, the candidate user is recommended to the target user. In this way, based on the social probability, not only can user recommendation be carried out in a targeted manner, but also by performing sentiment analysis on the candidate user, users with positive sentiment can be recommended to the target user, bringing positive and beneficial effects to the target user.
[0086] Exemplarily, the historical dissemination content can include social media posts, comments, blog articles, etc., and can be specifically set according to actual needs.
[0087] Exemplarily, when performing sentiment analysis on candidate users based on historical dissemination content, machine learning or deep learning methods can be used, such as Naive Bayes, Support Vector Machine (SVM), Recurrent Neural Network, Long Short-Term Memory (LSTM), etc. These methods can automatically learn text features and perform sentiment analysis to predict the sentiment analysis results.
[0088] Based on any of the above embodiments, exemplarily, in the embodiment of the present application, after recommending candidate users to the target user, for the first candidate user added as a friend by the target user, the behavior information of the first candidate user can also be monitored; and based on the registration information and behavior information of the first candidate user, the security level value of the first candidate user is re-determined; in the case where the security level value of the first candidate user is less than or equal to the preset level value, a prompt message is output to the terminal corresponding to the target user, and the prompt message is used to remind that the first candidate user is a risk user. In this way, by outputting the prompt message to the target user, unnecessary interference caused by the first candidate user to the target user can be avoided.
[0089] It can be understood that in the embodiment of the present application, the specific implementation of re-determining the security level value of the first candidate user based on the registration information and behavior information of the first candidate user is similar to the specific implementation of determining the security level value of the candidate user based on the registration information and behavior information in S102 above. For details, please refer to the above related description. Here, the embodiment of the present application will not be elaborated further.
[0090] Exemplarily, the prompt message can be a text prompt message or a graphic and text prompt message, etc., and can be specifically set according to actual needs.
[0091] Next, the user recommendation device provided by the present application will be described. The user recommendation device described below can be correspondingly referred to the user recommendation method described above.
[0092] Figure 3 It is a schematic structural diagram of a user recommendation device provided by an embodiment of the present application. Exemplarily, please refer to Figure 3 As shown, the user recommendation device 30 may include:
[0093] A screening unit 301, configured to screen multiple candidate users from multiple registered users based on the user profile of the target user when the target user is a secure user;
[0094] The first processing unit 302 is configured to determine a security level value for each candidate user based on the registration information and behavior information of the candidate user; and in the case where the security level value is greater than a preset level value, predict a social probability that the candidate user socializes with the target user based on the historical social behavior of the candidate user.
[0095] The recommendation unit 303 is configured to recommend the candidate user to the target user in the case where the social probability corresponding to the candidate user is greater than a preset probability.
[0096] Exemplarily, in an embodiment of the present application, the user profile includes location information, interest preferences, social circles, and behavior information. The screening unit 301 is configured to screen a plurality of candidate users from a plurality of registered users based on the user profile of the target user, including:
[0097] Respectively determine weights corresponding to the location information, interest preferences, social circles, and behavior information of the target user; wherein the weights are used to represent the attention degrees of the target user to the location information, interest preferences, social circles, and behavior information respectively.
[0098] For each registered user, respectively determine a first matching degree between the location information of the target user and the location information of the registered user, a second matching degree between the interest preferences of the target user and the interest preferences of the registered user, a third matching degree between the social circle of the target user and the social circle of the registered user, and a fourth matching degree between the behavior information of the target user and the behavior information of the registered user; and determine a target matching degree between the target user and the registered user based on the first matching degree, the second matching degree, the third matching degree, the fourth matching degree, and the weights corresponding to the location information, interest preferences, social circles, and behavior information respectively.
[0099] Screen the plurality of candidate users from the plurality of registered users based on the target matching degrees between the target user and each of the registered users.
[0100] Exemplarily, in an embodiment of the present application, the user recommendation device 30 further includes:
[0101] An acquisition unit, configured to acquire the location information, interest preferences, social circles, and behavior information of a first candidate user who is already added as a friend by the target user among all the recommended candidate users.
[0102] A matching unit, configured to match the location information, interest preferences, social circles, and behavior information of the target user with the location information, interest preferences, social circles, and behavior information of the first candidate user respectively to obtain a first matching result.
[0103] An adjustment unit, configured to adjust the weights corresponding to the location information, interest preferences, and social circles of the target user based on the first matching result.
[0104] Exemplarily, in an embodiment of the present application, the adjustment unit is configured to adjust the weights corresponding to the location information, interest preferences, social circles, and behavior information of the target user based on the first matching result, including:
[0105] When at least one of the target information in the location information, interest preferences, social circles, and behavior information of the target user fails to match successfully in the first matching result, match the target information of the target user with the target information of the second candidate user to obtain a second matching result; wherein, the second candidate user is a candidate user that the target user has not added as a friend among all the recommended candidate users;
[0106] When the second matching result indicates a successful match, reduce the weights of each piece of the target information, and increase the weights of the other information except the target information in the location information, interest preferences, social circles, and behavior information of the target user.
[0107] Exemplarily, in an embodiment of the present application, the recommendation unit 303 is configured to recommend the candidate user to the target user when the social probability corresponding to the candidate user is greater than a preset probability, including:
[0108] When the social probability corresponding to the candidate user is greater than a preset probability, obtain the historical dissemination content of the candidate user;
[0109] Perform sentiment analysis on the candidate user based on the historical dissemination content to obtain a sentiment analysis result; wherein, the sentiment analysis result includes positive sentiment, neutral sentiment, or negative sentiment;
[0110] When the sentiment analysis result is positive sentiment, recommend the candidate user to the target user.
[0111] Exemplarily, in an embodiment of the present application, the user recommendation device 30 further includes:
[0112] A monitoring unit, configured to monitor the behavior information of the first candidate user added as a friend by the target user;
[0113] A second processing unit, configured to re-determine the security level value of the first candidate user based on the registration information and behavior information of the first candidate user;
[0114] An output unit, configured to output a prompt message to a terminal corresponding to the target user when a security level value of the first candidate user is less than or equal to the preset level value, where the prompt message is used to remind that the first candidate user is a risky user.
[0115] Exemplarily, in an embodiment of the present application, the first processing unit 302 is configured to determine a security level value of the candidate user based on registration information and behavior information of the candidate user, including:
[0116] Inputting the registration information and behavior information of the candidate user into a security level model, and obtaining the security level value output by the security level model;
[0117] The security level model is obtained by training an initial security level model based on sample registration information, sample behavior information of multiple sample users, and security level value labels respectively corresponding to the multiple sample users.
[0118] The user recommendation device 30 provided in an embodiment of the present application can execute the technical solution of the user recommendation method in any of the above embodiments. Its implementation principle and beneficial effects are similar to those of the user recommendation method. For details, refer to the implementation principle and beneficial effects of the user recommendation method, which will not be elaborated here.
[0119] Figure 4 FIG. is a schematic physical structure diagram of an electronic device provided in an embodiment of the present application. As Figure 4 shown, the electronic device may include: a processor 410, a communication interface 420, a memory 430, and a communication bus 440. The processor 410, the communication interface 420, and the memory 430 communicate with each other through the communication bus 440. The processor 410 may call logic instructions in the memory 430 to execute the user recommendation method, and the method includes: when the target user is a secure user, screening multiple candidate users from multiple registered users based on a user profile of the target user; for each candidate user, determining a security level value of the candidate user based on registration information and behavior information of the candidate user; when the security level value is greater than a preset level value, predicting a social probability that the candidate user socializes with the target user based on historical social behavior of the candidate user; and recommending the candidate user to the target user when the social probability corresponding to the candidate user is greater than a preset probability.
[0120] In addition, when the logical instructions in the above-mentioned memory 430 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 this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. The foregoing storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.
[0121] On the other hand, this application also provides a computer program product. The computer program product includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the user recommendation method provided by the above-mentioned various methods. The method includes: when the target user is a secure user, based on the user profile of the target user, screening multiple candidate users from multiple registered users; for each candidate user, based on the registration information and behavior information of the candidate user, determining the security level value of the candidate user; when the security level value is greater than a preset level value, based on the historical social behavior of the candidate user, predicting the social probability of the candidate user socializing with the target user; when the social probability corresponding to the candidate user is greater than a preset probability, recommending the candidate user to the target user.
[0122] In yet another aspect, this application also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is implemented to execute the user recommendation method provided by the above-mentioned various methods. The method includes: when the target user is a secure user, based on the user profile of the target user, screening multiple candidate users from multiple registered users; for each candidate user, based on the registration information and behavior information of the candidate user, determining the security level value of the candidate user; when the security level value is greater than a preset level value, based on the historical social behavior of the candidate user, predicting the social probability of the candidate user socializing with the target user; when the social probability corresponding to the candidate user is greater than a preset probability, recommending the candidate user to the target user.
[0123] The device embodiments described above are merely illustrative. 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 to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative work.
[0124] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0125] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of the present application.
Claims
1. A user recommendation method, characterized in that, Including: When the target user is a secure user, screening multiple candidate users from multiple registered users based on the user profile of the target user; For each candidate user, determining a security level value of the candidate user based on the registration information and behavior information of the candidate user; When the security level value is greater than a preset level value, predicting a social probability that the candidate user will socialize with the target user based on the historical social behavior of the candidate user; When the social probability corresponding to the candidate user is greater than a preset probability, recommending the candidate user to the target user.
2. The user recommendation method according to claim 1, wherein The user profile includes location information, interest preferences, social circles, and behavior information. Screening multiple candidate users from multiple registered users based on the user profile of the target user includes: Respectively determining weights corresponding to the location information, interest preferences, social circles, and behavior information of the target user; wherein, the weights are used to represent the attention degrees of the target user to the location information, interest preferences, social circles, and behavior information respectively; For each registered user, respectively determining a first matching degree between the location information of the target user and the location information of the registered user, a second matching degree between the interest preferences of the target user and the interest preferences of the registered user, a third matching degree between the social circle of the target user and the social circle of the registered user, and a fourth matching degree between the behavior information of the target user and the behavior information of the registered user; and determining a target matching degree between the target user and the registered user based on the first matching degree, the second matching degree, the third matching degree, the fourth matching degree, and the weights corresponding to the location information, interest preferences, social circles, and behavior information respectively; Screening the multiple candidate users from the multiple registered users based on the target matching degrees between the target user and each registered user.
3. The user recommendation method according to claim 2, wherein The method further includes: Obtaining the location information, interest preferences, social circles, and behavior information of a first candidate user who is already added as a friend by the target user among all the recommended candidate users; Matching the location information, interest preferences, social circles, and behavior information of the target user with the location information, interest preferences, social circles, and behavior information of the first candidate user respectively to obtain a first matching result; Adjusting the weights corresponding to the location information, interest preferences, social circles, and behavior information of the target user based on the first matching result.
4. The user recommendation method according to claim 3, wherein Adjusting the weights corresponding to the location information, interest preferences, social circles, and behavior information of the target user based on the first matching result includes: When the first matching result indicates that at least one target information among the location information, interest preferences, social circles, and behavior information of the target user fails to match successfully, matching the target information of the target user with the target information of a second candidate user to obtain a second matching result; wherein, the second candidate user is a candidate user who is not added as a friend by the target user among all the recommended candidate users; In the case that the second matching result indicates successful matching, reduce the weights of the target information and increase the weights of other information except the target information among the location information, interest preferences, social circles, and behavioral information of the target user.
5. The user recommendation method according to claim 1 or 2, wherein The step of recommending the candidate user to the target user when the social probability corresponding to the candidate user is greater than a preset probability includes: When the social probability corresponding to the candidate user is greater than a preset probability, obtain the historical dissemination content of the candidate user; Perform sentiment analysis on the candidate user based on the historical dissemination content to obtain a sentiment analysis result; wherein, the sentiment analysis result includes positive sentiment, neutral sentiment, or negative sentiment; When the sentiment analysis result is positive sentiment, recommend the candidate user to the target user.
6. The user recommendation method according to claim 1 or 2, characterized in that The method further includes: Monitor the behavioral information of the first candidate user who has been added as a friend by the target user; Based on the registration information and behavioral information of the first candidate user, re-determine the security level value of the first candidate user; When the security level value of the first candidate user is less than or equal to the preset level value, output a prompt message to the terminal corresponding to the target user, where the prompt message is used to remind that the first candidate user is a risky user.
7. The user recommendation method according to claim 1 or 2, wherein The step of determining the security level value of the candidate user based on the registration information and behavioral information of the candidate user includes: Input the registration information and behavioral information of the candidate user into a security level model to obtain the security level value output by the security level model; Wherein, the security level model is obtained by training an initial security level model based on the sample registration information, sample behavioral information of multiple sample users, and the security level value labels corresponding to the multiple sample users respectively.
8. A user recommendation device, characterized in that, It includes: A screening unit, configured to screen multiple candidate users from multiple registered users based on the user profile of the target user when the target user is a secure user; A first processing unit, configured to determine the security level value of each candidate user based on the registration information and behavioral information of the candidate user; When the security level value is greater than a preset level value, predict the social probability of the candidate user and the target user for social interaction based on the historical social behavior of the candidate user; A recommendation unit, configured to recommend the candidate user to the target user when the social probability corresponding to the candidate user is greater than a preset probability.
9. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the user recommendation method according to any one of claims 1 to 7.
10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the user recommendation method according to any one of claims 1 to 7.
11. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the user recommendation method according to any one of claims 1 to 7.