Method, device, equipment and product for determining user matching degree

Abstract

The invention relates to a method, a device, equipment and a product for determining a user matching degree. The method comprises the following steps: determining a reaching probability that a first user and a second user have an interactive behavior; the method further includes determining a transition probability of the first user for a third user, wherein the first user, the second user, and the third user are different types of users. In addition, the method further comprises the step of determining the matching degree between the second user and the third user based on the reach probability and the conversion probability.

Description

Technical Field

[0001] The present disclosure relates to the field of computers, and more particularly to a method, apparatus, device, and product for determining user matching. Background Art

[0002] In the multimedia material platform service system, there are many types of users, including users who have information promotion needs and want to publish information, users who have creative capabilities and dissemination channels and can assist in publishing information, and users who pay attention to various types of information and can consume information.

[0003] The platform can bring these users together by leveraging its integration and docking functions, so that users who assist in publishing information can reach cooperation with users who publish information on the platform. In the process of cooperation, users who assist in publishing information can provide materials for the information published by the users who publish information, and attract more users who consume information for the users who publish information, thus forming a virtuous cycle of information in the platform service. Summary of the invention

[0004] In a first aspect of an embodiment of the present disclosure, a method for determining user matching is provided. The method includes determining a reach probability that a first user has an interactive behavior with a second user. The method also includes determining a conversion probability of the first user for a third user, wherein the first user, the second user, and the third user are different types of users. In addition, the method also includes determining a matching degree between the second user and the third user based on the reach probability and the conversion probability.

[0005] In a second aspect of an embodiment of the present disclosure, a device for determining user matching is provided. The device includes a reach probability determination module configured to determine a reach probability of an interactive behavior between a first user and a second user. The device includes a conversion probability determination module configured to determine a conversion probability of a first user for a third user, wherein the first user, the second user, and the third user are different types of users. In addition, the device also includes a matching degree determination module configured to determine a matching degree between the second user and the third user based on the reach probability and the conversion probability.

[0006] In a third aspect of the embodiments of the present disclosure, an electronic device is provided, which includes a processor and a memory coupled to the processor, wherein the memory has instructions stored therein, and when the instructions are executed by the processor, the electronic device executes the method according to the first aspect.

[0007] In a fourth aspect of an embodiment of the present disclosure, a computer program product is provided, on which computer executable instructions are stored, wherein the computer executable instructions are executed by a processor to implement the method of the first aspect.

[0008] This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to limit the scope of the claimed subject matter. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, wherein:

[0010] Figure 1 A schematic diagram illustrating an example environment in which various embodiments of the present disclosure may be implemented;

[0011] Figure 2 A flow chart showing a method for determining user matching according to some embodiments of the present disclosure is shown;

[0012] Figure 3 A schematic diagram showing an example process of matching users based on user matching degrees according to some embodiments of the present disclosure;

[0013] Figure 4 A schematic diagram showing an example structure of a combined decision tree model according to some embodiments of the present disclosure;

[0014] Figure 5 A schematic diagram showing construction of a combined decision tree model and application of the combined decision tree model according to some embodiments of the present disclosure is shown;

[0015] Figure 6 A schematic diagram showing an example structure of a double tower model according to some embodiments of the present disclosure;

[0016] Figure 7 A block diagram showing an apparatus for determining user matching according to some embodiments of the present disclosure; and

[0017] Figure 8 A block diagram of a device capable of implementing various embodiments of the present disclosure is shown. DETAILED DESCRIPTION

[0018] It is understood that all user-related data involved in this technical solution should be obtained and used after the user's authorization. This means that in this technical solution, if the user's personal information needs to be used, the user's explicit consent and authorization are required before obtaining this data, otherwise the relevant data will not be collected and used. It should also be understood that when implementing this technical solution, relevant laws and regulations should be strictly observed during the collection, use and storage of data, and necessary technologies and measures should be taken to protect the user's data security and ensure the safe use of data.

[0019] It is understandable that before using the technical solutions disclosed in the embodiments of the present disclosure, the types, scope of use, usage scenarios, etc. of the personal information involved in the present disclosure should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.

[0020] For example, when receiving an active request from a user, a prompt message is sent to the user to clearly prompt the user that the operation requested to be performed will require obtaining and using the user's personal information. Thus, the user can autonomously choose whether to provide personal information to software or hardware such as an electronic device, application, server, or storage medium that performs the operation of the technical solution of the present disclosure according to the prompt message.

[0021] As an optional but non-limiting implementation, in response to receiving an active request from the user, the prompt information may be sent to the user in the form of a pop-up window, in which the prompt information may be presented in text form. In addition, the pop-up window may also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0022] It is understandable that the above notification and the process of obtaining user authorization are merely illustrative and do not constitute a limitation on the implementation of the present disclosure. Other methods that meet the relevant laws and regulations may also be applied to the implementation of the present disclosure.

[0023] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0024] In the description of the embodiments of the present disclosure, the term "including" and similar terms should be understood as open inclusion, that is, "including but not limited to". The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc. may refer to different or the same objects, unless explicitly stated. Other explicit and implicit definitions may also be included below.

[0025] As mentioned above, the platform can play its integration and docking functions, bring together various types of users, and allow users with information promotion needs and users who have creative capabilities and communication channels to assist in information release to reach cooperation, so that users of other types of consumption can directly consume information. In the relevant technology, the consumer user group corresponding to the user with information promotion needs is first determined, and then the consumer user group corresponding to the user who assists in releasing information is determined, and then the overlap between the two consumer user groups is determined to determine whether the user with information promotion needs and the user who has creative capabilities and communication channels to assist in information release are matched. However, in the process of determining the consumer group users corresponding to both parties, this method can usually only select the corresponding consumer groups from a few characteristics, flattening the complex consumer group portraits, which is too simple and one-sided, and cannot determine the consumer groups that are truly highly compatible with both parties, resulting in inaccurate matching between the two parties. At the same time, this method also increases the difficulty of matching between the two parties, affecting the efficiency and effectiveness of cooperation, thereby reducing user experience.

[0026] To this end, the present disclosure proposes a method for determining user matching, which accurately calculates the matching degree of both parties, thereby improving the effectiveness and efficiency of cooperation between the two parties. In a scenario involving three different types of users, namely, a first user, a second user, and a third user, the probability of contact between the first user and the second user for interactive behavior, as well as the conversion probability of the first user for the third user, are first determined respectively, and then based on these two probabilities, the matching degree between the second user and the third user is comprehensively calculated. Through this method, the accuracy of analyzing the association and adaptation between different types of users can be improved, so that based on the comparison of the precise matching degree between these user combinations, the third user who best matches the second user can be found, or the second user who best matches the third user can be found, thereby improving the feasibility and effectiveness of user combinations in activities such as cooperative promotion.

[0027] Figure 1 1 is a schematic diagram of an example environment 100 in which various embodiments of the present disclosure may be implemented. Figure 1As shown, the first user acts as a bridge between the second user and the third user, connecting the second user and the third user. In some embodiments, the second user 110 may be a user with creative ability and dissemination channels, who can assist in publishing and promoting information. In some embodiments, the third user 130 may be a user who has information promotion needs and wants to publish information. In some embodiments, the first user (such as the first user 120-1, the first user 120-2, and the first user 120-3, etc.) may be a user who pays attention to various types of information and can consume information.

[0028] refer to Figure 1 In order to find a suitable third user for the second user to realize the subsequent information promotion service, the matching degree of the second user and the third user can be determined with the help of the bridge role of the first user. In some embodiments, the contact probability (such as a1, a2, an, etc.) of the second user 110 and the first user (such as the first user 120-1, the first user 120-2 and the first user 120-n, etc.) can be calculated first. The contact probability refers to the possibility that a user (or information sender) can successfully contact another user (or target recipient) and trigger interactive behavior in a specific environment or system (such as a social platform, information dissemination network, etc.). In some embodiments, the first user is a user who has interacted with the second user at least once, and the interactive behavior may include likes, comments, shares, favorites, homepage clicks, etc. In some embodiments, the contact probability between the second user and the first user can be calculated with the help of a machine learning model (such as an XGboost model).

[0029] Continue to refer Figure 1 Similarly, the conversion probability (such as b1, b2, bn, etc.) of the first user (such as the first user 120-1, the first user 120-2, and the first user 120-n, etc.) to the third user 130 can also be calculated. The conversion probability is an indicator that measures the possibility of a user (usually referred to as a source user) completing a specific conversion behavior and being associated with another user (target user) in a specific scenario. In some embodiments, the conversion rate of the first user to the third user can be calculated with the help of a conversion model. For example, the conversion model can be a PVR (Page-Visit-Revenue) conversion model, which is a model for measuring and analyzing the conversion of user behavior on digital platforms such as websites or mobile applications. It mainly focuses on the complete process of users starting from visiting a page, going through a series of behaviors, and finally generating revenue. In some embodiments, the conversion probability can be calculated using a double-tower structure conversion model.

[0030] In some embodiments, after calculating the reach probability between the second user 110 and the first user and the conversion probability of the first user to the third user 130, the matching degree between the second user and the third user can be determined based on these two probabilities, so that a second user matching the third user can be determined, and a third user matching the second user can also be determined.

[0031] In this way, the association and adaptation between different types of users can be analyzed, and the accuracy of user matching can be improved. Based on the comparison of the precise matching degrees between these user combinations, the third user who best matches the second user can be found, or the second user who best matches the third user can be found, thereby improving the feasibility and effectiveness of user combinations in activities such as cooperative promotions.

[0032] The following will combine Figures 2 to 8 The method according to the embodiment of the present disclosure is described in detail. For ease of understanding, the specific data mentioned in the following description are exemplary and are not intended to limit the scope of protection of the present disclosure. It is understood that the embodiments described below may also include additional actions not shown and / or the actions shown may be omitted, and the scope of the present disclosure is not limited in this respect.

[0033] Figure 2 A flow chart of a method 200 for determining a user matching degree according to some embodiments of the present disclosure is shown. The method 200 may be executed by a device for determining a user matching degree, which may be, for example, an independent device or system. The device may be implemented in software and / or hardware. Next, the method 200 is schematically described by taking the device for determining a user matching degree as an example of an execution subject. The method 200 includes a frame 202, a frame 204, and a frame 206.

[0034] like Figure 2 As shown, in block 202, the probability of reaching the first user and the second user having an interactive behavior is determined. Figure 1In order to find a suitable third user for the second user to realize the subsequent information promotion service, the matching degree of the second user and the third user can be determined with the help of the bridge role of the first user. In some embodiments, the contact probability (such as a1, a2, an, etc.) of the second user 110 and the first user (such as the first user 120-1, the first user 120-2 and the first user 120-n, etc.) can be calculated first. The contact probability refers to the possibility that a user (or information sender) can successfully contact another user (or target recipient) and trigger interactive behavior in a specific environment or system (such as a social platform, information dissemination network, etc.). In some embodiments, the first user is a user who has interacted with the second user at least once, and the interactive behavior may include likes, comments, shares, favorites, homepage clicks, etc. In some embodiments, the contact probability between the second user and the first user can be calculated with the help of a machine learning model (such as an XGboost model).

[0035] In block 204, a conversion probability of the first user to the third user is determined, wherein the first user, the second user, and the third user are different types of users. Figure 1 Similarly, the conversion probability (such as b1, b2, bn, etc.) of the first user (such as the first user 120-1, the first user 120-2, and the first user 120-n, etc.) to the third user 130 can also be calculated. The conversion probability is an indicator that measures the possibility of a user (usually referred to as a source user) completing a specific conversion behavior and being associated with another user (target user) in a specific scenario. In some embodiments, the conversion rate of the first user to the third user can be calculated with the help of a conversion model. For example, the conversion model can be a PVR conversion model, which is a model for measuring and analyzing the conversion of user behavior on digital platforms such as websites or mobile applications. It mainly focuses on the complete process of users starting from visiting a page, going through a series of behaviors, and finally generating revenue. In some embodiments, the conversion probability can be calculated using a double-tower structure conversion model.

[0036] In block 206, the matching degree between the second user and the third user is determined based on the reach probability and the conversion probability. Figure 1 In some embodiments, after calculating the reach probability between the second user 110 and the first user and the conversion probability of the first user to the third user 130, the matching degree between the second user and the third user can be determined based on these two probabilities ai and bi, so that a second user matching the third user can be determined, and a third user matching the second user can also be determined.

[0037] Through this method, the association and adaptation between different types of users can be analyzed, and the accuracy of user matching can be improved. Based on the comparison of the precise matching degrees between these user combinations, the third user who best matches the second user can be found, or the second user who best matches the third user can be found, thereby improving the feasibility and effectiveness of user combinations in activities such as cooperation.

[0038] Figure 3 FIG300 is a schematic diagram showing an example process of matching users based on user matching according to some embodiments of the present disclosure. Figure 3 At 310, a machine learning model is used to calculate the reach probability, that is, the machine learning model is used to calculate the reach probability of the second user to the first user.

[0039] In some embodiments, the machine learning model can be a self-supervised learning model. In some embodiments, the machine learning model can be a combined decision tree model (such as an XGboost model). The XGboost model is a machine learning algorithm based on a gradient boosting framework and can be viewed as a combined decision tree model. Figure 4 to describe. Figure 4 FIG. 4 is a schematic diagram showing an example structure 400 of a combined decision tree model according to some embodiments of the present disclosure. Figure 4 As shown in the figure, from a structural point of view, XGboost is composed of multiple decision trees. Its core idea is to fit the data by continuously adding decision trees. Each newly added decision tree is to make up for the shortcomings of the previous decision tree combination in prediction.

[0040] When calculating the probability of the first user reaching the second user, the interactive behavior data 401 (play volume, play time, likes, comments, shares, favorites, and homepage clicks, etc.) can be used as the input feature vector, and then each decision tree in the decision tree combination model can be used in turn.

[0041] In some embodiments, the interactive behavior data may be operation data such as likes, comments, shares, and favorites of the media data posted by the second user by the first user within a certain period of time, or the interactive behavior data may also be data such as the playback duration and number of playbacks of the media video posted by the first user when the second user clicks to play.

[0042] Actions such as liking, commenting, sharing, and collecting are positive feedback from the first user to the content posted by the second user. Liking indicates that the first user recognizes the content; commenting indicates that the first user not only watched the content but also participated in the discussion, which reflects the first user's high participation; sharing means that the first user thinks the content is valuable and wants to recommend it to others; collecting indicates that the first user may want to view the content again. These actions all indicate that the first user pays more attention to the second user. The number of video views reflects the exposure and attractiveness of the second user's content. A higher number of views means that the second user's content is seen by more users and may attract a wider user group. The length of play indicates the first user's involvement in the second user's video content. If the first user watches a video for a long time, it means that the content is attractive to the first user, and the first user may also be more interested in the second user.

[0043] refer to Figure 4 , taking decision tree 402 as an example, the input feature vector will start from the root node and pass down along the corresponding branch according to the feature splitting conditions on the node, such as whether the playback volume is greater than a certain threshold, until it reaches the leaf node. At this time, the leaf node will output a score or value, that is, weight 403. After that, the scores output by all decision trees in the combined decision tree model at the leaf nodes will be combined to determine the reach probability.

[0044] In some embodiments, a weighted summation method can be used to obtain the maximum probability. Assume that there are three decision trees, and the output scores for a first user are 0.2, 0.3, and 0.1 respectively. After weighted summation (assuming the weights are the same), the result is 0.2, and the initial estimate of the probability of reaching is 0.2. In order to make the output of the model more consistent with the probability form, this initial estimate can be mapped to a definition form that is more consistent with the probability distribution.

[0045] In the training phase, the decision tree of the combined decision tree model is constructed to fit the data pattern and learn the mapping relationship between input features such as the interaction behavior between the first user and the second user and the output label such as whether the user is reached. Each decision tree has a series of nodes and branches. Figure 5 to describe. Figure 5 A schematic diagram of constructing a combined decision tree model and applying the combined decision tree model 500 according to some embodiments of the present disclosure is shown.

[0046] like Figure 5As shown, in order to use the combined decision tree model to calculate the probability of contact between users, the combined decision tree model needs to be trained first. Features 502 for training can be extracted from the interactive behavior data 501 used for training. In some embodiments, these collected interactive data for user training are legally obtained from a first user who has interacted with a second user. In some embodiments, 70% of the collected data can be used to train the combined decision tree model, 20% of the data can be used to verify and adjust the tree depth, learning rate and other parameters of the combined decision tree, and 10% can be used to evaluate the final performance of the combined decision tree model.

[0047] In some embodiments, the interaction data of the first user with the second user within a certain period of time can be divided into positive samples and negative samples for training the combined decision tree model. In some embodiments, the positive samples indicate that the first user has interacted with the second user within a period of time, while the negative samples indicate that the first user has not interacted with the second user within a period of time.

[0048] In some embodiments, after obtaining the interaction data of the first user with respect to the second user, features 502 with positive and negative labels 503 can be extracted from these data, so that a combined decision tree model can be trained based on these features 502 with labels 503 to obtain a constructed training model 504, and thus a combined decision tree model 505 for reasoning can be obtained.

[0049] In some embodiments, after the training of the combined decision tree model is completed, the combined decision tree model can be used for reasoning applications. Figure 5 As shown, data can be applied at 506, and features 507 can be extracted from the data to be inferred and input into the combined decision tree model 505 of the application state, so that the combined decision tree model can apply the interactive behavior data of the first user with respect to the second user to predict the probability 508 of the second user reaching the first user.

[0050] By calculating the reach probability with the help of a machine learning model and taking into account multiple interactive behavior characteristics, the reach probability of the second user to the first user can be determined more accurately.

[0051] return Figure 3 At 320, a PVR conversion model is used to calculate the conversion probability. In some embodiments, a dual-tower PVR conversion model can be used to calculate the conversion probability, which can more accurately predict the conversion probability of the first user to the third user.

[0052] The following will be combined Figure 6 to describe. Figure 6A schematic diagram of an example structure 600 of a dual-tower model according to some embodiments of the present disclosure is shown. In some embodiments, the conversion probability of a first user for a third user can be obtained through a PVR conversion model of a dual-tower structure. The dual-tower model is a deep learning model architecture that is mainly composed of two independent neural network "towers", one for processing user-related features (user tower) and the other for processing item (such as service, content) related features (item tower). The outputs of these two towers are usually vector representations of fixed dimensions.

[0053] refer to Figure 6 , the first user-side feature 601 can be input into the user tower corresponding to the dual-tower model, where Figure 6 The feature vector layer 602 and the deep neural network 603 can form a user tower. The first user side feature 601 (various attribute portrait information of the first user including basic information, behavior habits, consumption history, etc.) can be converted into a low-dimensional vector form through the feature vector layer 602, and then after being processed by the deep neural network 703, the first user side feature vector 604 can be output. It can be understood that the collection and application of these data are all legal and agreed by the user.

[0054] Similarly, refer to Figure 6 , the third user-side feature 611 can be input into the item tower corresponding to the double-tower model, where Figure 6 The feature vector layer 612 and the deep neural network 613 can form an item tower. The third user side feature 611 (various industry information of the third user, including basic information, service information, promotion information, etc.), also called item feature, can be converted into a low-dimensional vector form through the feature vector layer 612, and then after being processed by the deep neural network 713, the third user side feature vector 614 can be output.

[0055] In some embodiments, the similarity of the first user side feature vector 604 and the third user side feature vector 614 output by the user tower can be calculated, and then the similarity can be converted into the conversion probability 620 of the first user to the third user through a specific mapping function. In some embodiments, the similarity can be calculated by cosine similarity, dot product and other methods. Through this dual-tower structure PVR conversion model, the conversion rate of the first user to the third user can be accurately calculated.

[0056] In some embodiments, the PVR conversion model of the dual-tower structure can also be trained by positive and negative samples. Among them, the positive sample indicates that the first user and the third user have a conversion relationship, while the negative sample indicates that there is no conversion relationship between the first user and the third user. By using this method of training the dual-tower model with positive and negative samples, the discrimination ability of the dual-tower model can be improved, so that the dual-tower model can more accurately predict the conversion probability. At the same time, the generalization ability of the dual-tower model is also enhanced, avoiding the overfitting problem caused by relying only on positive samples, so that the dual-tower model can see a more comprehensive data distribution, so that it can predict effective conversion efficiency.

[0057] return Figure 3 After the reach probability of the second user for the first user and the conversion probability of the first user for the third user are calculated, the matching degree between the third user and the second user can be calculated at 330, so that the user set can be determined based on the calculated matching degree at 340. In some embodiments, the user set can be a third user set required by the second user or a third user set required by the second user.

[0058] In some embodiments, in the process of calculating the matching degree, the total concentration can be calculated first at 332, including the total concentration of the second user to the third user and the total concentration of the third user to the second user. Afterwards, the matching degree between the users can be determined at 334 based on the total concentration and the respective matching coefficients. In some embodiments, if a set of matching second users is determined for the third user, the total concentration of the second user to the third user can be determined first, and then the matching degree of the second user to the third user can be determined based on the total concentration and the matching coefficient of the second user. In some embodiments, if a set of matching third users is determined for the second user, the total concentration of the third user to the second user can be determined first, and then the matching degree of the third user to the second user can be determined based on the total concentration and the matching coefficient of the third user.

[0059] In some embodiments, the total concentration of the second user for the third user (or the total concentration of the second user for the third user) can be calculated based on formula (1), which is as follows:

[0060] The second user targets the third user (or the second user targets the third user) Among them, a i is the probability of the second user reaching the first user, b i is the conversion probability of the first user to the third user.

[0061] In some embodiments, the matching degree of the second user to the third user (or the matching degree of the second user to the third user) can be calculated based on formula (2), which is as follows:

[0062] The matching degree of the second user to the third user (or the second user to the third user) = total concentration / matching coefficient (2) wherein the matching coefficient is related to whether the third user is screened for the second user or the second user is screened for the third user.

[0063] In some embodiments, when the matching degree of the second user to the third user is determined, a set of second users for the third user can be obtained, and the second users in the set can be sorted so that the third user can determine the second user that matches him / her.

[0064] Similarly, in some embodiments, when the matching degree of the third user to the second user is determined, a set of third users for the second user can be obtained, and the third users in the set can be sorted so that the second user can determine the third user that matches him / her.

[0065] In this way, the association and adaptation between different types of users can be analyzed, and based on the comparison of the matching degrees between these user combinations, the third user who best matches the second user can be found, or the second user who best matches the third user can be found, thereby improving the feasibility and effectiveness of user combinations in activities such as cooperative promotion.

[0066] Figure 7 FIG. 7 is a block diagram of an apparatus 700 for determining a user matching degree according to some embodiments of the present disclosure. Figure 7 As shown, the device 700 includes a reach probability determination module 702, which is configured to determine the reach probability of the first user and the second user having an interactive behavior. The device 700 also includes a conversion probability determination module 704, which is configured to determine the conversion probability of the first user for the third user, wherein the first user, the second user and the third user are different types of users. In addition, the device 700 also includes a matching degree determination module 706, which is configured to determine the matching degree between the second user and the third user based on the reach probability and the conversion probability.

[0067] Figure 8 8 is a block diagram of a device 800 capable of implementing various embodiments of the present disclosure. Figure 8As shown, the device 800 includes a central processing unit (CPU) and / or a graphics processing unit (GPU) 801, which can perform various appropriate actions and processes according to computer program instructions stored in a read-only memory (ROM) 802 or loaded from a storage unit 808 to a random access memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the device 800 can also be stored. The CPU / GPU 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804. Although not shown in FIG. Figure 8 As shown in FIG. 8 , device 800 may also include a co-processor.

[0068] A number of components in the device 800 are connected to the I / O interface 805, including: an input unit 806, such as a keyboard, a mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a disk, an optical disk, etc.; and a communication unit 809, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 809 allows the device 800 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0069] The various methods or processes described above may be performed by the CPU / GPU 801. For example, in some embodiments, the methods may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 800 via the ROM 802 and / or the communication unit 809. When the computer program is loaded into the RAM 803 and executed by the CPU / GPU 801, one or more steps or actions in the methods or processes described above may be performed.

[0070] In some embodiments, the methods and processes described above may be implemented as a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present disclosure.

[0071] Computer readable storage medium can be a tangible device that can hold and store instructions used by an instruction execution device. Computer readable storage medium can be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific examples (non-exhaustive list) of computer readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, for example, a punch card or a convex structure in a groove on which instructions are stored, and any suitable combination thereof. The computer readable storage medium used here is not interpreted as a transient signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagated by a waveguide or other transmission medium (for example, a light pulse by an optical fiber cable), or an electrical signal transmitted by a wire.

[0072] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0073] The computer program instructions for performing the disclosed operation may be assembler instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, programming languages ​​including object-oriented programming languages, and conventional procedural programming languages. Computer-readable program instructions may be executed completely on a user's computer, partially on a user's computer, executed as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., utilizing an Internet service provider to connect via the Internet). In certain embodiments, by utilizing the state information of a computer-readable program instruction to customize an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit may execute a computer-readable program instruction, thereby realizing various aspects of the present disclosure.

[0074] These computer-readable program instructions can be provided to a processing unit of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processing unit of the computer or other programmable data processing device, a device that implements the functions / actions specified in one or more boxes in the flowchart and / or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing device, and / or other equipment to work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0075] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0076] The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the equipment, method and computer program product according to multiple embodiments of the present disclosure. In this regard, each frame in the flow chart or block diagram can represent a part of a module, program segment or instruction, and a part of the module, program segment or instruction includes one or more executable instructions for realizing the specified logical function. In some alternative implementations, the function marked in the frame can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous frames can actually be executed substantially in parallel, and they can also be executed in the opposite order sometimes, depending on the functions involved. It should also be noted that each frame in the block diagram and / or flow chart, and the combination of frames in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0077] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The choice of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

[0078] Some example implementations of the present disclosure are listed below.

[0079] Example 1. A method for determining user matching, comprising:

[0080] Determine the probability of interaction between the first user and the second user;

[0081] determining a conversion probability of a first user to a third user, wherein the first user, the second user, and the third user are different types of users; and

[0082] Based on the reach probability and the conversion probability, a matching degree between the second user and the third user is determined.

[0083] Example 2. The method according to Example 1, wherein determining the probability of reaching the first user and the second user having an interactive behavior comprises:

[0084] determining a first feature of the first user with respect to the second user, the first feature indicating interaction between the first user and the second user within a period; and

[0085] The reach probability is determined by a first probability model based on the first feature.

[0086] Example 3. The method according to any one of Examples 1-2, wherein determining the reach probability based on the first feature by a first probability model comprises:

[0087] Based on the feature splitting condition of the node, the first feature is transmitted downward from the root node of each decision tree of the first probability model until a leaf node is reached; and

[0088] The weight values ​​output by the leaf nodes of each decision tree are combined to determine the reach probability.

[0089] Example 4. According to the method described in any one of Examples 1-3, the interaction situation includes at least one of the likes, comments, shares, favorites, plays, play duration, and play counts of the media content posted by the second user by the first user.

[0090] Example 5. The method according to any one of Examples 1-4, further comprising:

[0091] Determine a positive sample and a negative sample for the first probability model, the positive sample indicating that the first user has reached the target user within the cycle time, and the negative sample indicating that the first user has not reached the target user within the cycle time; and

[0092] The first probability model is trained based on the positive samples and the negative samples.

[0093] Example 6. According to the method described in any one of Examples 1-5, the first probability model is a supervised learning model.

[0094] Example 7. The method of any one of Examples 1-6, wherein determining the conversion probability of the first user to the third user comprises:

[0095] Determine a first user-side feature for the first user and a third user-side feature for the third user, wherein the first user-side feature indicates a portrait attribute of the first user, and the third user-side feature indicates an industry attribute of the third user; and

[0096] The conversion probability is determined by a second probability model based on the first user-side feature and the third user-side feature.

[0097] Example 8. According to the method of any one of Examples 1-7, the second probability model is a dual-tower model, wherein determining the conversion probability by the second probability model based on the first user-side feature and the third user-side feature includes:

[0098] The user tower determines a first vector representation based on the first user-side feature;

[0099] The item tower determines a second vector representation based on the third user-side feature; and

[0100] The conversion probability is determined based on a similarity between the first vector representation and the second vector representation.

[0101] Example 9. The method according to any one of Examples 1-8, further comprising:

[0102] Determining a positive sample and a negative sample for the second probability model, the positive sample indicating that a conversion relationship exists between the first user and the third user, and the negative sample indicating that no conversion relationship exists between the first user and the third user; and

[0103] The second probability model is trained based on the positive samples and the negative samples.

[0104] Example 10. The method according to any one of Examples 1-9, wherein determining the degree of match between the second user and the third user based on the reach probability and the conversion probability comprises:

[0105] Based on the reach probability and the conversion probability, determining a first set of second users related to the third user; and

[0106] In the first set, a second user matching the third user is determined.

[0107] Example 11. The method of any one of Examples 1-10, wherein in the first set, determining a second user that matches the third user comprises:

[0108] Based on the reach probability and the conversion probability, determining a total concentration of the second user to the third user, the total concentration indicating an ability of the second user to guide a plurality of the first users to have a conversion relationship with the third user;

[0109] determining a matching degree of the second user with respect to the third user based on the total concentration and a first matching coefficient, the first matching coefficient being determined based on the second user; and

[0110] Based on the matching degree, the second users in the first set are sorted in order to determine a second user matching the third user.

[0111] Example 12. The method according to any one of Examples 1-11, further comprising:

[0112] Based on the reach probability and the conversion probability, determining a second set of third users related to the second user; and

[0113] In the second set, a third user matching the second user is determined.

[0114] Example 13. The method of any one of Examples 1-12, wherein in the second set, determining a third user that matches the second user comprises:

[0115] Based on the contact probability and the conversion probability, determining a total concentration of the third user to the second user, the total concentration indicating an ability of the third user to guide a plurality of the first users to have a contact relationship with the second user;

[0116] determining a matching degree of the second user with respect to the third user based on the total concentration and a second matching coefficient, the second matching coefficient being determined based on the third user; and

[0117] Based on the matching degree, the third users in the second set are sorted in order to determine a third user matching the second user.

[0118] Example 14. An apparatus for determining a user matching degree, comprising:

[0119] A reach probability determination module, configured to determine a reach probability of an interactive behavior between the first user and the second user;

[0120] a conversion probability determination module, configured to determine a conversion probability of a first user with respect to a third user, wherein the first user, the second user, and the third user are users of different types; and

[0121] The matching degree determination module is configured to determine the matching degree between the second user and the third user based on the reach probability and the conversion probability.

[0122] Example 15. The apparatus of Example 14, wherein the reach probability determination module comprises:

[0123] A first determining module is configured to determine a first feature of the first user with respect to the second user, wherein the first feature indicates an interaction between the first user and the second user within a period; and

[0124] The second determination module is configured to determine the reach probability based on the first feature by a first probability model.

[0125] Example 16. The apparatus of any of Examples 14-15, wherein the second determining module comprises:

[0126] a feature transfer module, which transfers the first feature downward from the root node of each decision tree of the first probability model until it reaches a leaf node based on the feature splitting condition of the node; and

[0127] The third determination module is configured to combine the weight values ​​output by the leaf nodes of each decision tree to determine the reach probability.

[0128] Example 17. According to the device described in any one of Examples 14-16, the interaction situation includes at least one of the like, comment, share, favorite, play, play duration and play count of the media content posted by the second user by the first user.

[0129] Example 18. The apparatus of any one of Examples 14-17, further comprising:

[0130] a fourth determination module, configured to determine a positive sample and a negative sample for the first probability model, wherein the positive sample indicates that the first user has reached the target user within the cycle time, and the negative sample indicates that the first user has not reached the target user within the cycle time; and

[0131] The first training module is configured to train the first probability model based on the positive samples and the negative samples.

[0132] Example 19. In the apparatus according to any one of Examples 14-18, the first probability model is a supervised learning model.

[0133] Example 20. The apparatus of any one of Examples 14-19, wherein the conversion probability determination module comprises:

[0134] a fifth determination module, configured to determine a first user-side feature for the first user and a third user-side feature for the third user, wherein the first user-side feature indicates a portrait attribute of the first user, and the third user-side feature indicates an industry attribute of the third user; and

[0135] The sixth determination module is configured to determine the conversion probability based on the first user-side feature and the third user-side feature by a second probability model.

[0136] Example 21. The apparatus of any one of Examples 14-20, wherein the second probability model is a double tower model, wherein the sixth determination module comprises:

[0137] A first vector determination module is configured to determine a first vector representation based on the first user-side feature by a user tower;

[0138] A second vector determination module is configured to determine a second vector representation based on the third user-side feature by the item tower; and

[0139] A seventh determination module is configured to determine the conversion probability based on a similarity between the first vector representation and the second vector representation.

[0140] Example 22. The apparatus of any one of Examples 14-21, further comprising:

[0141] an eighth determination module, configured to determine a positive sample and a negative sample for the second probability model, wherein the positive sample indicates that a conversion relationship exists between the first user and the third user, and the negative sample indicates that no conversion relationship exists between the first user and the third user; and

[0142] The second training module is configured to train the second probability model based on the positive samples and the negative samples.

[0143] Example 23. The apparatus of any one of Examples 14-22, wherein the matching degree determination module comprises:

[0144] a ninth determination module, configured to determine a first set of second users related to the third user based on the reach probability and the conversion probability; and

[0145] The tenth determining module is configured to determine a second user matching the third user in the first set.

[0146] Example 24. The apparatus of any of Examples 14-23, wherein the tenth determining module comprises:

[0147] an eleventh determination module, configured to determine a total concentration of the second user to the third user based on the reach probability and the conversion probability, wherein the total concentration indicates an ability of the second user to guide a plurality of the first users to have a conversion relationship with the third user;

[0148] a twelfth determination module, configured to determine a matching degree of the second user related to the third user based on the total concentration and a first matching coefficient, wherein the first matching coefficient is determined based on the second user; and

[0149] A thirteenth determination module is configured to sort the second users in the first set in order based on the matching degree, and determine a second user matching the third user.

[0150] Example 25. The apparatus of any one of Examples 14-24, further comprising:

[0151] A fourteenth determining module is configured to determine a second set of third users related to the second user based on the reach probability and the conversion probability; and

[0152] A fifteenth determination module is configured to determine a third user matching the second user in the second set.

[0153] Example 26. The apparatus of any of Examples 14-25, wherein the eleventh determining module comprises:

[0154] A sixteenth determination module is configured to determine a total concentration of the third user to the second user based on the reach probability and the conversion probability, wherein the total concentration indicates an ability of the third user to guide a plurality of the first users to have a reach relationship with the second user;

[0155] A seventeenth determination module is configured to determine a matching degree of the second user related to the third user based on the total concentration and a second matching coefficient, wherein the second matching coefficient is determined based on the third user; and

[0156] An eighteenth determination module is configured to sort the third users in the second set in order based on the matching degree, and determine a third user matching the second user.

[0157] Example 27. An electronic device comprising:

[0158] Processor; and

[0159] A memory coupled to the processor, the memory having instructions stored therein, wherein when the instructions are executed by the processor, the electronic device performs actions, the actions comprising:

[0160] Determine the probability of interaction between the first user and the second user;

[0161] determining a conversion probability of a first user to a third user, wherein the first user, the second user, and the third user are different types of users; and

[0162] Based on the reach probability and the conversion probability, a matching degree between the second user and the third user is determined.

[0163] Example 28. The electronic device according to Example 27, wherein determining the contact probability that the first user and the second user have an interactive behavior comprises:

[0164] determining a first feature of the first user with respect to the second user, the first feature indicating interaction between the first user and the second user within a period; and

[0165] The reach probability is determined by a first probability model based on the first feature.

[0166] Example 29. The electronic device of any one of Examples 27-28, wherein determining the reach probability based on the first feature by a first probability model comprises:

[0167] Based on the feature splitting condition of the node, the first feature is transmitted downward from the root node of each decision tree of the first probability model until a leaf node is reached; and

[0168] The weight values ​​output by the leaf nodes of each decision tree are combined to determine the reach probability.

[0169] Example 30. According to the electronic device described in any one of Examples 27-29, the interaction situation includes at least one of the likes, comments, shares, favorites, plays, play duration, and play counts of the media content posted by the second user by the first user.

[0170] Example 31. The electronic device of any of Examples 27-30, further comprising:

[0171] Determine a positive sample and a negative sample for the first probability model, the positive sample indicating that the first user has reached the target user within the cycle time, and the negative sample indicating that the first user has not reached the target user within the cycle time; and

[0172] The first probability model is trained based on the positive samples and the negative samples.

[0173] Example 32. In the electronic device described in any one of Examples 27-31, the first probability model is a supervised learning model.

[0174] Example 33. An electronic device according to any one of Examples 27-32, wherein determining a conversion probability of the first user to a third user comprises:

[0175] Determine a first user-side feature for the first user and a third user-side feature for the third user, wherein the first user-side feature indicates a portrait attribute of the first user, and the third user-side feature indicates an industry attribute of the third user; and

[0176] The conversion probability is determined by a second probability model based on the first user-side feature and the third user-side feature.

[0177] Example 34. The electronic device according to any one of Examples 27-33, wherein the second probability model is a dual-tower model, wherein determining the conversion probability by the second probability model based on the first user-side feature and the third user-side feature comprises:

[0178] The user tower determines a first vector representation based on the first user-side feature;

[0179] The item tower determines a second vector representation based on the third user-side feature; and

[0180] The conversion probability is determined based on a similarity between the first vector representation and the second vector representation.

[0181] Example 35. The electronic device of any of Examples 27-34, further comprising:

[0182] Determining a positive sample and a negative sample for the second probability model, the positive sample indicating that a conversion relationship exists between the first user and the third user, and the negative sample indicating that no conversion relationship exists between the first user and the third user; and

[0183] The second probability model is trained based on the positive samples and the negative samples.

[0184] Example 36. The electronic device according to any one of Examples 27-35, wherein based on the reach probability and the conversion probability, determining the matching degree between the second user and the third user comprises:

[0185] Based on the reach probability and the conversion probability, determining a first set of second users related to the third user; and

[0186] In the first set, a second user matching the third user is determined.

[0187] Example 37. The electronic device of any of Examples 27-36, wherein determining, in the first set, a second user that matches the third user comprises:

[0188] Based on the reach probability and the conversion probability, determining a total concentration of the second user to the third user, the total concentration indicating an ability of the second user to guide a plurality of the first users to have a conversion relationship with the third user;

[0189] determining a matching degree of the second user with respect to the third user based on the total concentration and a first matching coefficient, the first matching coefficient being determined based on the second user; and

[0190] Based on the matching degree, the second users in the first set are sorted in order to determine a second user matching the third user.

[0191] Example 38. The electronic device of any of Examples 27-37, further comprising:

[0192] Based on the reach probability and the conversion probability, determining a second set of third users related to the second user; and

[0193] In the second set, a third user matching the second user is determined.

[0194] Example 39. The electronic device of any of Examples 27-38, wherein determining a third user matching the second user in the second set comprises:

[0195] Based on the contact probability and the conversion probability, determining a total concentration of the third user to the second user, the total concentration indicating an ability of the third user to guide a plurality of the first users to have a contact relationship with the second user;

[0196] determining a matching degree of the second user with respect to the third user based on the total concentration and a second matching coefficient, the second matching coefficient being determined based on the third user; and

[0197] Based on the matching degree, the third users in the second set are sorted in order to determine a third user matching the second user.

[0198] Example 40. A computer-readable storage medium having computer-executable instructions stored thereon, wherein the computer-executable instructions are executed by a processor to implement a method according to any one of Examples 1 to 13.

[0199] Example 41. A computer program product tangibly stored on a computer-readable medium and comprising computer-executable instructions that, when executed by a device, cause the device to perform a method according to any one of Examples 1 to 13.

[0200] Although the disclosure has been described in language specific to structural features and / or methodological logical actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Instead, the specific features and actions described above are merely example forms of implementing the claims.

Claims

1. A method for determining user matching, comprising: Determine the probability of interaction between the first user and the second user; determining a conversion probability of a first user to a third user, wherein the first user, the second user, and the third user are users of different types; as well as Based on the reach probability and the conversion probability, a matching degree between the second user and the third user is determined.

2. The method according to claim 1, wherein determining the contact probability that the first user and the second user have an interactive behavior comprises: determining a first feature of the first user with respect to the second user, the first feature indicating interaction between the first user and the second user within a period; as well as The reach probability is determined by a first probability model based on the first feature.

3. The method according to claim 2, wherein determining the reach probability based on the first feature by a first probability model comprises: Based on the feature splitting condition of the node, the first feature is transmitted downward from the root node of each decision tree of the first probability model until a leaf node is reached; as well as The weight values ​​output by the leaf nodes of each decision tree are combined to determine the reach probability.

4. According to the method of claim 2, the interaction situation includes at least one of the likes, comments, shares, favorites, plays, play duration and play counts of the media content posted by the second user by the first user.

5. The method according to claim 2, further comprising: Determine a positive sample and a negative sample for the first probability model, the positive sample indicating that the first user has reached the target user within the cycle time, and the negative sample indicating that the first user has not reached the target user within the cycle time; as well as The first probability model is trained based on the positive samples and the negative samples. The method according to claim 5 , wherein the first probability model is a supervised learning model.

7. The method according to claim 1, wherein determining the conversion probability of the first user to the third user comprises: Determine a first user-side feature for the first user and a third user-side feature for the third user, wherein the first user-side feature indicates a portrait attribute of the first user, and the third user-side feature indicates an industry attribute of the third user; and The conversion probability is determined by a second probability model based on the first user-side feature and the third user-side feature.

8. The method according to claim 7, wherein the second probability model is a double-tower model, wherein determining the conversion probability by the second probability model based on the first user-side feature and the third user-side feature comprises: The user tower determines a first vector representation based on the first user-side feature; The item tower determines a second vector representation based on the third user-side feature; as well as The conversion probability is determined based on a similarity between the first vector representation and the second vector representation.

9. The method according to claim 7, further comprising: Determining a positive sample and a negative sample for the second probability model, the positive sample indicating that a conversion relationship exists between the first user and the third user, and the negative sample indicating that no conversion relationship exists between the first user and the third user; and The second probability model is trained based on the positive samples and the negative samples.

10. The method according to claim 1, wherein determining the matching degree between the second user and the third user based on the reach probability and the conversion probability comprises: Based on the reach probability and the conversion probability, determining a first set of second users related to the third user; as well as In the first set, a second user matching the third user is determined.

11. The method according to claim 10, wherein in the first set, determining a second user matching the third user comprises: Based on the reach probability and the conversion probability, determining a total concentration of the second user to the third user, the total concentration indicating an ability of the second user to guide a plurality of the first users to have a conversion relationship with the third user; determining a matching degree of the second user associated with the third user based on the total concentration and a first matching coefficient, wherein the first matching coefficient is determined based on the second user; as well as Based on the matching degree, the second users in the first set are sorted in order to determine a second user matching the third user.

12. The method according to claim 10, further comprising: Based on the reach probability and the conversion probability, determining a second set of third users related to the second user; as well as In the second set, a third user matching the second user is determined.

13. The method according to claim 12, wherein in the second set, determining a third user matching the second user comprises: Based on the contact probability and the conversion probability, determining a total concentration of the third user to the second user, the total concentration indicating an ability of the third user to guide a plurality of the first users to have a contact relationship with the second user; determining a matching degree of the second user related to the third user based on the total concentration and a second matching coefficient, wherein the second matching coefficient is determined based on the third user; as well as Based on the matching degree, the third users in the second set are sorted in order to determine a third user matching the second user.

14. A device for determining a user matching degree, comprising: A reach probability determination module, configured to determine a reach probability of an interactive behavior between the first user and the second user; a conversion probability determination module, configured to determine a conversion probability of a first user with respect to a third user, wherein the first user, the second user and the third user are users of different types; as well as The matching degree determination module is configured to determine the matching degree between the second user and the third user based on the reach probability and the conversion probability.

15. An electronic device, comprising: processor; as well as A memory coupled to the processor, the memory having instructions stored therein, wherein when the instructions are executed by the processor, the electronic device executes the method according to any one of claims 1 to 13.

16. A computer program product comprising computer executable instructions, wherein the computer executable instructions are executed by a processor to implement the method according to any one of claims 1 to 13.

Images