A method, apparatus, device, and medium for user screening

Abstract

The application belongs to the technical field of data processing, and mainly relates to machine learning technology in artificial intelligence, and discloses a user screening method, device, equipment and medium. The user screening method disclosed by the application comprises the following steps: determining the intimacy between each two users according to online social behavior data between the users, determining the influence of each user according to the intimacy between each two users, and screening target users whose influence meets a set condition. In this way, the influence of users is determined through online social behavior data between the users, the accuracy of user influence determination is improved, and users with influence can be accurately screened.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to a user screening method and device, equipment and medium. BACKGROUND

[0002] With the development of Internet technology, the network has gradually become an indispensable part of people's daily life. Various network projects (such as network games) are also increasing.

[0003] In order to promote the development of network projects, it is usually necessary to screen out users with influence in the network project, so that other users of the network project can be maintained through the screened users.

[0004] Therefore, how to accurately screen out users with influence is a problem to be solved. SUMMARY

[0005] The embodiments of the present application provide a user screening method, device, equipment and medium, which can improve the accuracy of determining the influence of users when screening users.

[0006] In one aspect, a user screening method is provided, comprising:

[0007] obtaining online social behavior data between users;

[0008] determining the intimacy between each two users according to the online social behavior data between users;

[0009] determining the influence of each user according to the intimacy between each two users;

[0010] determining target users whose influence meets a set condition according to the influence of users.

[0011] In one aspect, a user screening device is provided, comprising:

[0012] an acquisition unit configured to obtain online social behavior data between users;

[0013] a first determination unit configured to determine the intimacy between each two users according to the online social behavior data between users;

[0014] a second determination unit configured to determine the influence of each user according to the intimacy between each two users;

[0015] a screening unit configured to screen out target users whose influence meets a set condition.

[0016] Preferably, the acquisition unit is configured to:

[0017] respectively obtaining a behavior value of each type of online social behavior between each two users.

[0018] Preferably, the first determining unit is configured to:

[0019] respectively for each two users, performing the following steps: normalizing each behavior value between the two users, and performing weighted summation on the normalized behavior values to obtain an intimacy between the two users.

[0020] Preferably, the second determining unit is configured to:

[0021] establishing an intimacy matrix according to the intimacy between each two users;

[0022] performing iterative processing on the intimacy matrix by using a preset ranking algorithm to obtain the influence of each user;

[0023] The ranking algorithm is configured to determine the influence of each user according to the intimacy between the user and each of the other users.

[0024] Preferably, the second determining unit is configured to:

[0025] respectively for each two users, performing the following steps: determining a transition probability value according to the intimacy between the two users and the sum of the intimacy between one of the two users and each user, the transition probability value being positively correlated with the intimacy and negatively correlated with the sum;

[0026] obtaining the intimacy matrix according to the determined transition probability values.

[0027] Preferably, the second determining unit is further configured to:

[0028] performing convergence correction processing on the intimacy matrix according to the obtained total number of users to obtain a corrected intimacy matrix;

[0029] wherein each element value of the corrected intimacy matrix is negatively correlated with the total number of users.

[0030] In one aspect, a control device is provided, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor performing the steps of the above-mentioned method for screening users when executing the program.

[0031] In one aspect, a computer readable storage medium is provided, which stores a computer program, the computer program being executable on a processor to implement the steps of the above-mentioned method for screening users.

[0032] The method, device, equipment and medium for user screening provided by the embodiments of the present application determine the intimacy between each two users according to the online social behavior data between the users, determine the influence of each user according to the intimacy between each two users, and screen target users whose influence meets the set condition from the users. In this way, the influence of each user is determined according to the online social behavior data between the users, the accuracy of influence determination and the accuracy of target user screening are improved, the target users with influence who can maintain other users can be screened out, and the purpose of maintaining the retention and payment activities of other users by the target users with influence can be achieved.

[0033] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application can be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings. BRIEF DESCRIPTION OF DRAWINGS

[0034] The accompanying drawings, which are included to provide a further understanding of the present application and are incorporated in and constitute a part of this application, illustrate embodiments of the present application and serve to explain the present application. In the drawings:

[0035] Figure 1 An application scenario in the embodiments of the present application is shown in the figure;

[0036] Figure 2 An implementation flowchart of the method for user screening in the embodiments of the present application is shown in the figure;

[0037] Figure 3 A detailed implementation flowchart of the method for user screening in the embodiments of the present application is shown in the figure;

[0038] Figure 4a An implementation flowchart of an application example of the method for user screening in the embodiments of the present application is shown in the figure;

[0039] Figure 4b An example diagram of the shunting test in the embodiments of the present application is shown in the figure;

[0040] Figure 5 An example diagram of the evaluation index in the embodiments of the present application is shown in the figure;

[0041] Figure 6 A structure schematic diagram of the device for user screening in the embodiments of the present application is shown in the figure;

[0042] Figure 7 A structure schematic diagram of the control equipment in the embodiments of the present application is shown in the figure. DETAILED DESCRIPTION

[0043] In order to make the purposes, technical solutions and beneficial effects of the present application clearer, the present application is further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0044] Firstly, some terms involved in the embodiments of the present application are described to facilitate understanding by those skilled in the art.

[0045] Terminal device: a device that can install various applications and can display objects provided by the installed applications, which can be mobile or fixed. For example, a mobile phone, a tablet computer, various wearable devices, a vehicle-mounted device, a personal digital assistant (PDA), a point of sales (POS), or other electronic devices capable of achieving the above functions.

[0046] Artificial intelligence (AI): is to use digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology of computer science, which tries to understand the essence of intelligence and produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines, so that machines have the functions of perception, reasoning and decision-making.

[0047] Artificial intelligence technology is a comprehensive discipline, involving a wide range of fields, both hardware and software technologies. Artificial intelligence basic technologies generally include technologies such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction system, mechatronics, etc. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology, and machine learning / deep learning, etc.

[0048] Machine Learning (ML) is a multi-disciplinary subject, involving probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory, etc. It is a specialized study of how computers simulate or implement human learning behavior to acquire new knowledge or skills, reorganize existing knowledge structure to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental approach to making computers intelligent. Its applications are widespread in various fields of artificial intelligence. Machine learning and deep learning usually include artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and rule-based learning.

[0049] PageRank algorithm: a method for ranking web pages by using simple hyperlinks to calculate the score of web pages, so as to rank the web pages. In this application, it is used to calculate the influence of users according to the intimacy between users.

[0050] Online social behavior data: the behavior value of various online social behavior types between users. The online social behavior types can be chatting, teaming up in games, sending gifts, and private messages, etc. The behavior value of each online social behavior type can be the number of chats, the number of teaming up, and the number of private messages, etc.

[0051] The design idea of the embodiments of the present application is introduced as follows.

[0052] With the development of Internet technology, the network has gradually become an indispensable part of people's daily life. Various network projects are also increasing.

[0053] For network projects with strong social attributes, users with influence in the network project can usually be screened out, and other users of the network project can be maintained through the screened users to promote the development of the network project. For example, the screened users can be key opinion leaders (KOLs), that is, people who have more and more accurate product information, are accepted or trusted by the relevant group, and have a great influence on the purchase behavior of the group.

[0054] Taking a network project as an example, in the traditional way, when screening users with influence, the following two methods are usually used:

[0055] The first method is to determine the influence of users according to the ranking of individual combat power, the ranking of guild copies, and the circle layer to which the guild war belongs, and to screen users with high influence.

[0056] The second method is to construct a supervised training sample and use a tree network model to determine the influence of users and screen out users with high influence.

[0057] However, the accuracy of the influence determined by the above two manners is low, the value of the users screened is not great, and the number of friends is low, which does not play a role in maintaining other users.

[0058] Obviously, there is no technical solution for accurately determining the influence of each user in the prior art, and therefore, there is an urgent need for a technical solution for screening users, which improves the accuracy of determining the influence of users.

[0059] In view of the above analysis and consideration, the embodiments of the present application provide a data processing solution, in which the intimacy between each two users is determined according to the online social behavior data between the users, the influence of each user is determined according to the intimacy between each two users, and the target user whose influence meets a set condition is screened from the users. In this way, the influence of each user is determined according to the online social behavior data between the users, the accuracy of determining the influence is improved, and the target user with influence who can maintain other users is screened, and the accuracy of screening the target user is improved.

[0060] To further illustrate the technical solutions provided by the embodiments of the present application, the following will be described in detail in combination with the drawings and specific embodiments. Although the embodiments of the present application provide the method operation steps as described in the following embodiments or shown in the drawings, more or fewer operation steps can be included in the method based on conventional or non-creative labor. The execution order of the steps is not limited to the execution order provided by the embodiments of the present application in the logical sense that there is no necessary causal relationship between the steps. The method can be executed in sequence or in parallel when the method is actually processed or executed by the device.

[0061] The present application can be applied to various network projects with strong social attributes, for example, network games and network interactive courses. In the embodiments of the present application, the method for screening users is described by taking a network game as an example.

[0062] Referring to Figure 1 Fig. 1 is a schematic diagram of an application scenario. The application scenario includes a plurality of terminal devices 110 and a control device 120. A game client 111 is installed in each terminal device 110, and the game client 111 is embedded with an instant messaging application. The control device 120 can be one or a group of servers for providing instant messaging services and game services.

[0063] Figure 1 In Fig. 1, four terminal devices 110 are taken as an example, and in actual application, the number of terminal devices 110 is not limited.

[0064] After the user logs in the game client 111, the user can chat, form a team, send gifts and perform other online social activities with other users through the control device 120.

[0065] The control device 120 obtains online social behavior data of each user in the network game, determines the intimacy between each two users according to the online social behavior data of each user, determines the influence of each user according to the intimacy between each two users, and further screens target users that meet a set condition.

[0066] Optionally, the set condition can be a specified number of users with the highest influence. In actual applications, the set condition and the specified number can be set according to actual application scenarios, which are not limited herein.

[0067] In the embodiments of the present application, the influence of each user is determined according to the online social behavior data between users, so that the target users with high influence can be accurately screened. In actual applications, the management personnel of the operator can establish contact with the screened target users through the housekeeper portal, internal messages and telephone, and improve the game experience of the target users by using gift packages, benefits, preferential services and privileges, so as to ensure the game retention of the target users, establish the trust relationship between the management personnel and the target users, and after the trust relationship and the personal retention of the target users are established, the target users can influence the operation of guild activities through their own influence, for example, the target users can organize guild battles, play instances and other forms to influence the activity and payment of other users, so as to promote the operation of the network project.

[0068] Referring to Figure 2 FIG. 1 shows an implementation flowchart of a user screening method provided by the present application. The specific process of the method is as follows:

[0069] Step 200: The control device obtains online social behavior data between each two users.

[0070] Specifically, the control device obtains the behavior value of each online social behavior type between each two users.

[0071] The control device can obtain the online social behavior data according to the communication request messages, communication response messages and team formation messages of the recorded terminal devices. The online social behavior data is the behavior value of multiple online social behavior types between users, and the online social behavior types can be chatting, game team formation, gift sending and private messaging.

[0072] For example, the online social behavior data between user a and user b is that the two users chat 10 times in a week, chat for a total of 2 hours in a week, and form a team 5 times in a week.

[0073] For example, the online social behavior data between user c and user d is that they chat 10 times in a day and send a gift once in a week.

[0074] In this way, the behavior values of various online social behavior types between users can be obtained, and the social range, social manner, social duration, and social frequency between users can be determined according to the behavior values of different online social behavior types between different users.

[0075] Step 201: The control device determines the intimacy between each two users according to the online social behavior data between the users.

[0076] Specifically, the control device performs the following steps for each two users:

[0077] Each behavior value between the two users is normalized, and the normalized behavior values are weighted and summed to obtain the intimacy between the two users.

[0078] That is, a corresponding weight is set for each online social behavior type in advance, each behavior value between each two users is normalized, and the sum of the product of the normalized behavior value between each two users and the corresponding weight is calculated to obtain the corresponding intimacy.

[0079] In actual applications, the weights corresponding to different online social behavior types can be set according to actual application scenarios, which are not limited herein.

[0080] In one implementation, when step 201 is performed, the following steps can be performed:

[0081] S2011: The control device establishes a corresponding sub-network between each user for each online social behavior type.

[0082] Each online social behavior type corresponds to a sub-network, and the sub-network is composed of nodes, edges composed of node connections, and behavior values on the edges.

[0083] The node represents a user and can be the identification information of the user. The identification information of each user is unique. The edge represents the social relationship between users and can be represented as (u i ,u j ), where u represents a user, i.e., a node, and i and j represent the serial numbers of the users, which are both positive integers. The behavior value on the edge represents a quantitative indicator of the social behavior of the online social behavior type.

[0084] For example, the network project is a game of "King of Kings", and each node is a player in the game of "King of Kings". Each player has a corresponding role and a role account, such as a guild leader and a defense minister, and the role account of each player is the identification information of the user and is unique. The online social behavior type is a game team type, and a game team sub-network of each player is established. If two players are in a team relationship, a line between the nodes corresponding to the two players is connected as an edge of the two nodes, and the behavior value is the number of teaming between the two users.

[0085] For another example, assuming that the online social behavior type is a chat type, a chat sub-network of each user is established. Each user social account is a node, and if two users are in a friend relationship, a line between the nodes corresponding to the two users is connected as an edge of the two nodes, and the behavior value is the number of chats or chat frequency between the two users.

[0086] S2012: The control device normalizes each behavior value in each sub-network respectively to obtain a normalized behavior value.

[0087] Specifically, the control device performs the following steps for each sub-network: normalizing each behavior value respectively to obtain a normalized behavior value, so that the normalized behavior value is within the range of [0, 1], and the normalized behavior value of each sub-network is represented by e ij , which represents the normalized behavior value between user u i and user u j .

[0088] Since the behavior values of different sub-networks use different quantitative indicators, the behavior values of each online social behavior type are standardized by normalization in order to calculate the intimacy subsequently.

[0089] S2013: The control device determines the intimacy between each two users according to the normalized behavior value in each sub-network.

[0090] Specifically, the control device identifies the same nodes in each sub-network, and merges the nodes corresponding to the same user in each sub-network into the same node. The normalized behavior values of each online social behavior type between each two nodes are weighted and summed to obtain the corresponding intimacy.

[0091] The intimacy between two users can be calculated by the following formula:

[0092] E ij = w1e 1 ij + w2e 2 ij + … + w N e Nij ;

[0093] Where E represents intimacy, i and j represent user numbers (positive integers), w represents weight, e represents behavior value, and N represents the number of subnetworks.

[0094] In this way, the intimacy between each pair of users can be determined based on data showing that users engage in different types of online social behavior.

[0095] Step 202: The control device determines the influence of each user based on the intimacy between each pair of users.

[0096] Specifically, when performing step 202, the following steps can be adopted:

[0097] S2021: The control device establishes an affinity matrix based on the affinity between each pair of users.

[0098] Specifically, the control device performs the following steps for each pair of users: determining the corresponding transition probability value based on the intimacy between the two users and the sum of the intimacy between one of the users and the other users, and obtaining an intimacy matrix based on the determined transition probability values.

[0099] Among them, the transfer probability value is positively correlated with intimacy and negatively correlated with the sum of the above.

[0100] In one implementation, an initial intimacy matrix is ​​formed by the intimacy levels between users, and the initial intimacy matrix is ​​processed by transition probability to obtain an intimacy matrix after transition probability processing.

[0101] The initial intimacy matrix consists of elements representing the intimacy level E between two users. ij When processing the transition probabilities of the initial affinity matrix, the following steps can be used:

[0102] For each pair of users, perform the following steps: Based on user u i and user u j Intimacy level E between them ij , and u i The sum of the intimacy levels with each user determines the corresponding transition probability value.

[0103] In one implementation, the transition probability value can be expressed using the following formula:

[0104]

[0105] Where M is the transition probability value, E represents the intimacy level, i and j represent the user's serial number (both are positive integers), and m is the number of users.

[0106] In this way, the closeness between users is processed by using the transition probability, and a closeness matrix is obtained to improve data stability.

[0107] In S2022, the control device performs convergence correction processing on the closeness matrix to obtain a corrected closeness matrix.

[0108] Specifically, the control device performs convergence correction processing on the closeness matrix according to the total number of users to obtain a corrected closeness matrix. Each element value of the corrected closeness matrix is negatively correlated with the total number of users.

[0109] In an implementation, the control device determines a correction value corresponding to each transition probability value according to a product of a first setting parameter and the transition probability value, and a ratio of a second setting parameter and the total number of users, and obtains a corrected closeness matrix composed of the correction values.

[0110] The first setting parameter and the second setting parameter are both in the range of [0, 1], and the sum of the first setting parameter and the second setting parameter is 1. For example, a can be 0.85, and b can be 0.15.

[0111] In actual application, the first setting parameter and the second setting parameter can be set according to actual application scenarios, which are not limited here.

[0112] The correction value can be determined by the following formula:

[0113] P ij = a M ij + b / m

[0114] P ij is a correction value of the closeness between user u i and user u j , i and j represent the serial numbers of the users, a is the first setting parameter, b is the second setting parameter, M is the transition probability value, and m is the number of users.

[0115] In this way, the closeness matrix is processed by using the convergence correction processing, so that the data convergence is improved when the closeness matrix is iteratively processed in the subsequent steps.

[0116] In S2023, the control device iteratively processes the closeness matrix by using a preset ranking algorithm to obtain the influence of each user.

[0117] Specifically, the ranking algorithm is used to calculate the influence of the user according to the closeness between the users.

[0118] Optionally, the ranking algorithm can use the PageRank algorithm.

[0119] In an implementation, the influence determination can employ the following formula:

[0120] X k+1 =p k X0;

[0121] wherein X k+1 is an m-dimensional vector composed of the influence of each user, p is a corrected closeness matrix, k is the number of iterations, and X0is a set m-dimensional vector. The i-th element in X k+1 is the influence of the i-th user.

[0122] In actual applications, X0may be set according to the actual application scenario, which is not limited herein.

[0123] In this way, the influence of each user can be determined respectively.

[0124] Step 203: The control device screens out target users whose influence satisfies a set condition.

[0125] Specifically, the control device screens out a specified number of users with the highest influence from among all users as target users.

[0126] In actual applications, the specified number can be set according to the actual application scenario, for example, the specified number can be 5.

[0127] Referring to FIG. 7, a detailed implementation flowchart of a user screening method provided by the present application is shown. Figure 3

[0128] The specific flow of the method is as follows:

[0129] Step 300: The control device respectively acquires the behavior value of each online social behavior type between each two users.

[0130] Specifically, when step 300 is performed, refer to the detailed steps in step 200 above.

[0131] Step 301: The control device respectively establishes a corresponding sub-network between each user for each online social behavior type.

[0132] Specifically, when step 301 is performed, refer to the detailed steps in step 201 above.

[0133] Step 302: The control device respectively normalizes each behavior value in each sub-network to obtain a normalized behavior value.

[0134] Specifically, when step 302 is performed, refer to the detailed steps in step 201 above.

[0135] ​Step 303: The control device determines the intimacy between each two users according to the normalized behavior value in each sub-network.

[0136] Specifically, when step 303 is performed, refer to step 201 for details.

[0137] Step 304: The control device establishes an intimacy matrix according to the intimacy between each two users.

[0138] Specifically, when step 304 is performed, refer to step 202 for details.

[0139] Step 305: The control device performs convergence correction processing on the intimacy matrix to obtain a corrected intimacy matrix.

[0140] Specifically, when step 305 is performed, refer to step 202 for details.

[0141] Step 306: The control device performs iteration processing on the intimacy matrix using a preset ranking algorithm to obtain the influence of each user.

[0142] Specifically, when step 306 is performed, refer to step 202 for details.

[0143] Step 307: The control device screens out target users whose influence meets a set condition.

[0144] Specifically, when step 307 is performed, refer to step 203 for details.

[0145] Referring to Figure 4a FIG. 3 shows an embodiment of the implementation flowchart of the user screening application. The specific process of the method is as follows:

[0146] Step 400: The control device establishes a sub-network corresponding to each type of online social behavior according to the online social behavior data of each user.

[0147] Specifically, when step 400 is performed, refer to step S2011 for details.

[0148] Step 401: The control device normalizes each behavior value in each sub-network to obtain a normalized behavior value.

[0149] Specifically, when step 401 is performed, refer to step S2012 for details.

[0150] Step 402: The control device identifies the same nodes in each sub-network.

[0151] Step 403: The control device merges the nodes corresponding to the same user in each sub-network into the same node.

[0152] Step 404: The control device performs a weighted sum of the normalized behavior values ​​of each online social behavior type between every two nodes to obtain the corresponding intimacy level.

[0153] Specifically, when performing steps 402-404, please refer to step S2013 above for detailed steps.

[0154] Step 405: The control device establishes an intimacy matrix based on the intimacy between each pair of users.

[0155] Step 406: The control device uses the PageRank algorithm to iteratively process the affinity matrix to obtain the influence of each user.

[0156] Step 407: Control the device to filter out target users whose influence meets the set conditions.

[0157] In this embodiment, the intimacy level of each user is determined based on online social behavior data between users, and a probability transfer process is performed on the intimacy level of each user to improve the stability of the data. Furthermore, a convergence correction process is performed on the intimacy matrix obtained after the probability transfer process to avoid the problem of non-convergence during iterative processing of the intimacy matrix. This solves the problem of bidirectional flow in social relationships. By iterating the intimacy matrix multiple times, the influence of users is accurately determined, thereby accurately screening out high-influence target users.

[0158] See Figure 4b The diagram illustrates an example of a traffic splitting test. Multiple comparison combinations are set up; comparison combination 1, comparison combination 2, and comparison combination 3 are used as examples. Each comparison combination consists of a control group and an experimental group. In the control group, target users are selected from 240 users using a traditional user screening method. In the experimental group, target users are selected from 240 users using the user screening method described in this application. The traditional methods used in the control groups differ across comparison combinations, while the user screening method described in this application is used in all experimental groups.

[0159] See Figure 5 The image shown is an example of an evaluation metric. The evaluation metrics include total spending by friends, average spending per friend, number of active friends, and average number of active days per friend. These metrics are used to evaluate the target users selected from both the control group and the experimental group in each comparison combination.

[0160] Referring to Table 1, which is an example of a beneficial effect table. The evaluation index data of the target users screened in the control group and the experimental group in September is obtained respectively. The average game time of friends in September is 100, the average payment amount of friends is 10, the number of friend active users is 10, and the average active days per person is 10. In the experimental group, the average game time of friends in September is 105, the average payment amount of friends is 12, the number of friend active users is 13, and the average active days per person is 11. Compared with the control group, the average game time of friends in the experimental group is increased by 5%, the average payment amount of friends is increased by 20%, the number of friend active users is increased by 30%, and the average active days per person is increased by 10%.

[0161] Table 1.

[0162]

[0163] Based on the same inventive concept, the embodiments of the present application also provide a user screening device. Since the principles of the above device and equipment for solving problems are similar to those of a user screening method, the implementation of the above device can refer to the implementation of the method, and the repeated parts will not be described again.

[0164] As Figure 6 shown, it is a structure schematic diagram of a user screening device provided by the embodiments of the present application.

[0165] A user screening device comprises:

[0166] An acquisition unit 601 is configured to acquire online social behavior data between users.

[0167] A first determination unit 602 is configured to determine the intimacy between each two users according to the online social behavior data between users.

[0168] A second determination unit 603 is configured to determine the influence of each user according to the intimacy between each two users.

[0169] A screening unit 604 is configured to screen target users whose influence meets a set condition.

[0170] Preferably, the acquisition unit 601 is configured to:

[0171] Acquire the behavior value of each online social behavior type between each two users.

[0172] Preferably, the first determination unit 602 is configured to:

[0173] For each two users, the following steps are performed: normalize each behavior value between the two users, and weight and sum the normalized behavior values to obtain the intimacy between the two users.

[0174] Preferably, the second determining unit 603 is configured to:

[0175] build a closeness matrix according to the closeness between each two users;

[0176] obtain the influence of each user by iteratively processing the closeness matrix using a preset ranking algorithm.

[0177] The ranking algorithm is configured to determine the influence of each user according to the closeness between the user and each of the other users.

[0178] Preferably, the second determining unit 603 is configured to:

[0179] for each two users, determine a transition probability value according to the closeness between the two users and the sum of the closeness between one of the two users and each user, the transition probability value being positively correlated with the closeness and negatively correlated with the sum;

[0180] obtain the closeness matrix according to the determined transition probability values.

[0181] Preferably, the second determining unit 603 is further configured to:

[0182] perform convergence correction processing on the closeness matrix according to the total number of users to obtain a corrected closeness matrix;

[0183] wherein each element value of the corrected closeness matrix is negatively correlated with the total number of users.

[0184] In the method, device, equipment and medium for user screening provided by the embodiments, the closeness between each two users is determined according to the online social behavior data between the users, the influence of each user is determined according to the closeness between each two users, and the target user whose influence meets the set condition is screened out from the users. In this way, the influence of each user is determined according to the online social behavior data between the users, the accuracy of the determined influence is improved, and the target user with influence who can maintain other users is screened out, and the accuracy of the screened target user is improved.

[0185] Figure 7 A structural schematic diagram of a control device 7000 is shown. Referring to Figure 7 As shown, the control device 7000 includes a processor 7010, a memory 7020, a power supply 7030, a display unit 7040, and an input unit 7050.

[0186] The processor 7010 is a control center of the control device 7000, connects various components by using various interfaces and lines, and performs overall monitoring on the control device 7000 by running or executing software programs and / or data stored in the memory 7020 to perform various functions of the control device 7000.

[0187] In the embodiments of the present application, the processor 7010 executes the method for user screening provided by the embodiments shown in the Figure 2

[0188] Optionally, the processor 7010 can include one or more processing units; preferably, the processor 7010 can integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, the user interface and the application, and the modem processor mainly processes the wireless communication. It can be understood that the above-mentioned modem processor can also not be integrated into the processor 7010. In some embodiments, the processor, the memory, and the like can be implemented on a single chip, and in some embodiments, they can also be implemented on separate chips respectively.

[0189] The memory 7020 can mainly include a program storage area and a data storage area, wherein the program storage area can store the operating system, various applications and the like; the data storage area can store data created according to the use of the control device 7000 and the like. In addition, the memory 7020 can include a high-speed random access memory, and can also include a non-volatile memory, for example, at least one magnetic disk storage device, a flash memory device, or other volatile solid-state memory device and the like.

[0190] The control device 7000 further includes a power supply 7030 (such as a battery) for supplying power to various components, and the power supply can be logically connected to the processor 7010 through a power management system, so as to realize the functions of managing charging, discharging and power consumption and the like through the power management system.

[0191] The display unit 7040 can be used to display information input by a user or information provided to the user, and various menus of the control device 7000 and the like, and in the embodiments of the present application, is mainly used to display the display interface of each application in the control device 7000 and the text, pictures and the like objects displayed in the display interface. The display unit 7040 can include a display panel 7041. The display panel 7041 can be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED) and the like.

[0192] ​The input unit 7050 can be configured to receive information such as numbers or characters input by a user. The input unit 7050 can include a touch panel 7051 and other input devices 7052. The touch panel 7051, also called a touch screen, can collect touch operations of a user thereon or adjacent thereto, such as operations of the user using a finger, a stylus, or any suitable object or accessory on or adjacent to the touch panel 7051.

[0193] Specifically, the touch panel 7051 can detect a touch operation of a user, and detect signals caused by the touch operation, convert the signals into touch coordinates, send the touch coordinates to the processor 7010, and receive commands from the processor 7010 and execute the commands. In addition, the touch panel 7051 can be implemented in various types such as a resistive type, a capacitive type, an infrared type, and a surface acoustic wave type. The other input devices 7052 can include, but are not limited to, one or more of a physical keyboard, function keys (such as a volume control key, a power on / off key, and the like), a trackball, a mouse, a joystick, and the like.

[0194] Of course, the touch panel 7051 can cover the display panel 7041, and when the touch panel 7051 detects a touch operation thereon or adjacent thereto, the touch panel 7051 transmits the touch operation to the processor 7010 to determine a type of the touch event, and then the processor 7010 provides corresponding visual output on the display panel 7041 according to the type of the touch event. Although in the above description, the touch panel 7051 and the display panel 7041 are implemented as two independent components to control the input and output functions of the control device 7000, in some embodiments, the touch panel 7051 and the display panel 7041 can be integrated to control the input and output functions of the control device 7000. Figure 7

[0195] The control device 7000 can further include one or more sensors, such as a pressure sensor, a gravity acceleration sensor, a proximity light sensor, and the like. Of course, according to the needs in specific applications, the above control device 7000 can further include a camera and other components, since these components are not the components mainly used in the embodiments of the present application, they are not shown in the above description and will not be described in detail. Figure 7

[0196] Those skilled in the art can understand that the above description of the control device is merely an example of the control device, and does not constitute a limitation on the control device, and the control device can include more or fewer components than those shown in the figure, or combine some components, or different components. Figure 7

[0197] The embodiments of the present application also provide a computer readable storage medium, which stores a computer program. The computer program is executed by a processor to implement the method for user screening in any of the above method embodiments.

[0198] ​​​Those skilled in the art can clearly understand the implementation of the various embodiments by means of software plus a general hardware platform from the above description of the embodiments, and of course, the various embodiments can also be implemented by hardware. Based on such an understanding, the above technical solutions, essentially or in other words, the part that contributes to the related art, can be embodied in the form of a software product. The computer software product can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, and the like, and includes a plurality of instructions to cause a control device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of the various embodiments or some parts of the embodiments.

[0199] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for some technical features thereof; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A method of user screening, characterized by, The method comprises the following steps: respectively acquiring behavior values of each online social behavior type between each two users in a network game; the online social behavior type is one of chatting, game teaming, gifting and private messaging, each user is a player in the network game, and the behavior values of different online social behavior types between different users can determine the social range, social manner, social duration and social frequency between the users; for each online social behavior type, a corresponding sub-network between each two users is established, wherein one online social behavior type corresponds to one sub-network, each sub-network comprises nodes, edges formed by node connection, and behavior values of the corresponding online social behavior type, and each node corresponds to a role account of a user in the network game; each behavior value in each sub-network is normalized to obtain normalized behavior values; intimacy between each two users is determined according to the normalized behavior values in each sub-network and weights corresponding to each online social behavior type respectively; an initial intimacy matrix is formed by the intimacy between the users, and a transition probability processed intimacy matrix is obtained by performing transition probability processing on the initial intimacy matrix; a correction value corresponding to each element value in the transition probability processed intimacy matrix is determined according to a product between the first set parameter and each element value in the transition probability processed intimacy matrix and a ratio between the second set parameter and the total number of users, and a corrected intimacy matrix composed of the correction values is obtained; an influence of each user is obtained by performing iterative processing on the corrected intimacy matrix by using a preset ranking algorithm, and the ranking algorithm is used to determine the influence of a corresponding user according to the intimacy between each user and other users. target users whose influence meets a set condition are screened out.

2. The method of claim 1, wherein, The intimacy between each two users is determined according to the normalized behavior values in each sub-network, which comprises the following steps: for each two users, the following steps are performed: weighted sum of the normalized behavior values is performed to obtain the intimacy between the two users.

3. The method of claim 1, wherein, The initial intimacy matrix is formed by the intimacy between the users, and the transition probability processed intimacy matrix is obtained by performing transition probability processing on the initial intimacy matrix, which comprises the following steps: for each two users, the following steps are performed: according to the intimacy between the two users and a sum of the intimacy between the two users and the intimacy between one of the two users and each user, a corresponding transition probability value is determined, the transition probability value is positively correlated with the intimacy and negatively correlated with the sum; The transition probability processed intimacy matrix is obtained according to the determined transition probability values.

4. The method of claim 1 or 3, wherein, Each element value in the corrected intimacy matrix is negatively correlated with the total number of users.

5. An apparatus for user screening, characterized by The method comprises the following steps: an acquisition unit is configured to acquire behavior values of each online social behavior type between each two users in a network game; The online social behavior type is one of chatting, teaming up for a game, gifting and private messaging, each user is a player in the network game, and the behavior value of different online social behavior types between different users can determine the social range, social manner, social duration and social frequency between users. The first determining unit is configured to establish a corresponding sub-network between each two users for each online social behavior type, wherein one online social behavior type corresponds to one sub-network, each sub-network includes nodes, node connection edges and behavior values of the corresponding online social behavior type, each node corresponds to a role account of a user in the network game, each behavior value in each sub-network is normalized to obtain a normalized behavior value, and the closeness between each two users is determined according to the normalized behavior value in each sub-network and a weight corresponding to each online social behavior type. The second determining unit is configured to form an initial closeness matrix by the closeness between users, perform transition probability processing on the initial closeness matrix to obtain a closeness matrix after transition probability processing, determine a correction value corresponding to each element value in the closeness matrix after transition probability processing according to a product between the first setting parameter and the element value and a ratio between the second setting parameter and the total number of users, obtain a corrected closeness matrix composed of the correction values, and perform iterative processing on the corrected closeness matrix by using a preset ranking algorithm to obtain the influence of each user, wherein the ranking algorithm is configured to determine the influence of each user according to the closeness between the user and other users. The screening unit is configured to screen out target users whose influence meets a set condition.

6. A control device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to realize the steps of the method in any one of claims 1-4.

7. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the steps of the method in any one of claims 1-4.

Images