Browsing behavior information identification method, device, equipment and medium

Abstract

The embodiments of the present disclosure disclose a method, device, equipment and medium for identifying browsing behavior information. A specific implementation of the method includes: obtaining the historical browsing behavior data of the current user within a preset time period, wherein the historical browsing behavior data represents the behavior data of each page browsed by the user; generating a feature vector corresponding to the user based on the historical browsing behavior data; inputting the feature vector into the user browsing behavior associated node prediction model to obtain the user browsing behavior associated node prediction information, wherein the user browsing behavior associated node prediction information represents the predicted association relationship between the user and each node corresponding to the browsed page; in response to determining that the user browsing behavior associated node prediction information meets the preset conditions, the user information of the user is pushed to at least one associated node terminal. This implementation can identify whether there is a communication connection relationship between the user and the node terminal of the page.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of computer technology, and in particular to methods, devices, equipment, and media for identifying browsing behavior information. Background Art

[0002] When a user browses a page, it is often necessary to count the user's browsing data in order to push information to the user. Currently, when counting the user's browsing data, the method usually adopted is: determining the user's browsing time for the page and determining whether the user clicks other links in the page.

[0003] However, when using the above statistics on user browsing data, the following technical problems usually exist: only determining the browsing time of the page and the clicks on the link can only determine whether the user is interested in the page, but it is difficult to determine whether the user has a connection with the creator of the page (the page server), resulting in the information pushed to the user not meeting the needs and wasting push resources.

[0004] The above information disclosed in this Background section is only for enhancement of understanding of the background of the inventive concept and therefore it may contain information that does not form the prior art that is already known in this country to a person of ordinary skill in the art. Summary of the invention

[0005] The content of this disclosure is used to introduce concepts in a brief form, which will be described in detail in the detailed implementation section below. The content of this disclosure is not intended to identify the key features or essential features of the technical solution claimed for protection, nor is it intended to limit the scope of the technical solution claimed for protection.

[0006] Some embodiments of the present disclosure propose browsing behavior information identification methods, devices, electronic devices and computer-readable media to solve one or more of the technical problems mentioned in the above background technology section.

[0007] In a first aspect, some embodiments of the present disclosure provide a method for identifying browsing behavior information, the method comprising: obtaining historical browsing behavior data of a current user within a preset time period, wherein the historical browsing behavior data represents behavior data of each page browsed by the user; generating a feature vector corresponding to the user based on the historical browsing behavior data; inputting the feature vector into a pre-trained user browsing behavior associated node prediction model to obtain user browsing behavior associated node prediction information, wherein the user browsing behavior associated node prediction information represents the association relationship between the predicted user and each node corresponding to the browsed page; in response to determining that the user browsing behavior associated node prediction information meets a preset condition, pushing the user information of the user to at least one associated node terminal.

[0008] In a second aspect, some embodiments of the present disclosure provide a browsing behavior information identification device, the device comprising: an acquisition unit, configured to acquire historical browsing behavior data of a current user within a preset time period, wherein the historical browsing behavior data represents behavior data of each page browsed by the user; a generation unit, configured to generate a feature vector corresponding to the above-mentioned user based on the above-mentioned historical browsing behavior data; an input unit, configured to input the above-mentioned feature vector into a pre-trained user browsing behavior associated node prediction model to obtain user browsing behavior associated node prediction information, wherein the above-mentioned user browsing behavior associated node prediction information represents the association relationship between the predicted user and each node corresponding to the browsed page; a push unit, configured to push the user information of the above-mentioned user to at least one associated node terminal in response to determining that the above-mentioned user browsing behavior associated node prediction information meets a preset condition.

[0009] In a third aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation manner of the above-mentioned first aspect.

[0010] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method described in any implementation manner of the above-mentioned first aspect is implemented.

[0011] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the browsing behavior information identification method of some embodiments of the present disclosure, it is possible to identify whether there is a communication connection relationship between the user and the node terminal of the page, thereby improving the accuracy of the information pushed to the user and avoiding the waste of push resources. Specifically, the reason why the information pushed to the user does not meet the demand and wastes the push resources is that only by determining the browsing time of the page and the click of the link, it can only be determined whether the user is interested in the page, but it is difficult to determine whether the user has a connection with the creator of the page (page server), resulting in the information pushed to the user not meeting the demand and wasting push resources. Based on this, the browsing behavior information identification method of some embodiments of the present disclosure first obtains the historical browsing behavior data of the current user within a preset time period. Among them, the historical browsing behavior data represents the behavior data of each page browsed by the user. Therefore, the behavior data of each page can be used to provide data support for determining whether there is an association relationship between the user and each node corresponding to the browsed page. Then, according to the above-mentioned historical browsing behavior data, a feature vector corresponding to the above-mentioned user is generated. Therefore, it is convenient to determine the association relationship between the user and each node corresponding to the browsed page according to each feature in the feature vector. Then, the feature vector is input into a pre-trained user browsing behavior associated node prediction model to obtain user browsing behavior associated node prediction information. The user browsing behavior associated node prediction information indicates the association relationship between the predicted user and each node corresponding to the browsed page. Thus, the association relationship between the user and each node corresponding to the browsed page can be identified based on the pre-trained user browsing behavior associated node prediction model. Finally, in response to determining that the user browsing behavior associated node prediction information meets the preset conditions, the user information of the user is pushed to at least one associated node terminal. Thus, the user information can be pushed to the node terminal with the associated relationship, so that the node terminal can accurately push the information to the user, avoiding the waste of push resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] 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. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that components and elements are not necessarily drawn to scale.

[0013] Figure 1 is a flow chart of some embodiments of the browsing behavior information identification method according to the present disclosure;

[0014] Figure 2 is a schematic diagram of the structure of some embodiments of the browsing behavior information identification device according to the present disclosure;

[0015] Figure 3 It is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0016] 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 set forth herein. On the contrary, these embodiments are provided to provide 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.

[0017] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present disclosure can be combined with each other.

[0018] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0019] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0020] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0021] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0022] Figure 1 1 is a flow chart of some embodiments of the method for identifying browsing behavior information according to the present disclosure. 100 of some embodiments of the method for identifying browsing behavior information according to the present disclosure is shown. The method for identifying browsing behavior information comprises the following steps:

[0023] Step 101, obtaining historical browsing behavior data of the current user within a preset time period.

[0024] In some embodiments, the execution subject (such as a computing device) of the browsing behavior information identification method can obtain the historical browsing behavior data of the current user within a preset time period through a wired connection or a wireless connection. Among them, the historical browsing behavior data represents the behavior data of each page browsed by the user. The historical browsing behavior data includes: a custom event set. The custom event set may include: user identity data, behavior log, device information, geographic location data, network traffic data, login history data, permission and access control data, browsing behavior event set, third-party data, time series data, page usage data, log and event data and other custom events. For example, user identity data includes basic identity information such as the user's name, user name, email address, and phone number. The behavior log includes the user's login record, session activity, access time and frequency, etc. The device information may include information such as the device type, operating system, and browser used by the user. The geographic location data may include: the user's IP address and geographic location information. The network traffic data may include: the user's data transmission on the network, including incoming and outgoing data traffic, protocol usage, etc. The login history data may include: the user's login history, including successful and failed login attempts, and the login location. Permission and access control data include: user's permission level, accessed resources and files, access control policies, etc. Browsing behavior events in the browsing behavior event set may include: user click operations on a certain page, browsing time and page interaction, which can provide information about user interests and social networks. Third-party data may include: data from external data providers, such as credit scores, blacklist data, etc. Time series data may refer to time series data of user behavior, which can reveal periodic activities or abnormal behavior. Page usage data may include: user usage on each page, including the services used, frequency and duration, etc. Log and event data may include: abnormal events, alarms and error messages for each page. One custom event corresponds to one page.

[0025] Step 102: Generate a feature vector corresponding to the user based on the historical browsing behavior data.

[0026] In some embodiments, the execution entity may generate a feature vector corresponding to the user based on the historical browsing behavior data.

[0027] In practice, the execution entity may generate a feature vector corresponding to the user through the following steps:

[0028] The first step is to determine the custom event set using the following formula:

[0029] S={(A1,c1,t1),(A2,c2,t2)…,(A n ,cn ,t n )}.

[0030] Among them, S represents a custom event set, A1, A2, A n is a custom event set, c1, c2, c n Indicates the access time to the custom event, t1, t2, t n Indicates the dwell time (browsing duration) of a custom event.

[0031] The second step is to determine the total access time of the custom event set using the following formula:

[0032]

[0033] Where c is the total access time mentioned above, c i is the access time to the custom event, and n is the number of custom events in the custom event set.

[0034] The third step is to determine the total dwell time of the custom event set using the following formula:

[0035]

[0036] Where T is the total residence time mentioned above, t i The dwell time for custom events.

[0037] Step 4: Determine the proportion of each access time in the total access time using the following formula:

[0038]

[0039] in, It represents the proportion of the i-th access time in the total access time.

[0040] Step 5: Determine the proportion of each dwell time in the total dwell time using the following formula:

[0041]

[0042] in, is the proportion of the i-th residence time in the total residence time.

[0043] Step 6: Determine the user's interest in each custom event using the following formula:

[0044]

[0045] Among them, ω i is the i-th interest level.

[0046] In the seventh step, the custom event set, the total access time, the total residence time, the proportion of each access time in the total access time, the proportion of each residence time in the total residence time, and the user's interest in each custom event are combined into a feature vector corresponding to the user.

[0047] Therefore, the user's feature vector can be constructed by combining the user's historical behavior data. By analyzing and profiling the user's behavior, the association relationship between the user and the node terminal can be determined in a targeted manner.

[0048] Optionally, obtain a set of user historical browsing behavior information.

[0049] In some embodiments, the execution entity may obtain a user historical browsing behavior information set. The user historical browsing behavior information in the user historical browsing behavior information set includes a response tag, a node intervention tag, and a feature information set. The response tag may characterize the user's page click event. The node intervention tag may characterize the type of response made to the user's page click event. Page click events may include: icon click events, page exit click events, and page link click events. The node intervention tag may be an intervention tag of the node terminal for the user (e.g., feedback on the click event). The feature information set may be the average feature information corresponding to a user's historical browsing behavior, and may include: user identity data, behavior logs, device information, geographic location data, network traffic data, login history data, permissions and access control data, browsing behavior event sets, third-party data, time series data, page usage data, logs, and event data.

[0050] Optionally, based on the above user historical browsing behavior information set, a browsing response label classification network and a node intervention label classification network are trained.

[0051] In some embodiments, the above-mentioned execution subject can train the browsing response label classification network and the node intervention label classification network according to the above-mentioned user historical browsing behavior information set. Among them, the above-mentioned browsing response label classification network and the above-mentioned node intervention label classification network both correspond to feature sequences. Each feature in the feature sequence can be arranged in descending order of importance. In practice, first, the above-mentioned execution subject can use the response label included in each user historical browsing behavior information in the user historical browsing behavior information set as a sample label, and use the node intervention label and feature information set included in each user historical browsing behavior information in the user historical browsing behavior information set as the input model feature, and train the response label classification network according to the above-mentioned user historical browsing behavior information set. For example, the response label classification network can be LightGBM. There is no limitation on the model type of the response label classification network. Then, the various features used for judgment in the classification process of the response label classification network can be combined in sequence to obtain the feature sequence corresponding to the above-mentioned response label classification network. Secondly, the node intervention labels included in each user's historical browsing behavior information in the user's historical browsing behavior information set can be used as sample labels, and the response labels and feature information sets included in each user's historical browsing behavior information in the user's historical browsing behavior information set can be used as input features. The node intervention label classification network is trained according to the above user's historical browsing behavior information set. For example, the node intervention label classification network can be LightGBM. There is no limitation on the model type of the node intervention label classification network. Then, the various features used for judgment in the classification process of the node intervention label classification network can be combined in sequence to obtain a feature sequence corresponding to the above node intervention label classification network.

[0052] Optionally, a browsing behavior causal feature set is generated according to a feature sequence corresponding to the browsing response label classification network and a feature sequence corresponding to the node intervention label classification network.

[0053] In some embodiments, the execution entity may generate a browsing behavior causal feature set according to a feature sequence corresponding to the browsing response label classification network and a feature sequence corresponding to the node intervention label classification network.

[0054] In practice, the above execution entity can generate a browsing behavior causal feature set through the following steps:

[0055] In the first step, each feature that satisfies a preset feature condition in the feature sequence corresponding to the browsing response label classification network is determined as a first feature group, wherein the preset feature condition may be that the feature is a preset number of features arranged in front.

[0056] In the second step, each feature that meets the preset feature conditions in the feature sequence corresponding to the above-mentioned node intervention classification network is determined as a second feature group.

[0057] In the third step, the intersection of the first feature group and the second feature group is determined as the browsing behavior causal feature set.

[0058] Optionally, based on the browsing behavior causal feature set and the user historical browsing behavior information set, the initial user browsing behavior associated node prediction model is trained to obtain a trained user browsing behavior associated node prediction model.

[0059] In some embodiments, the execution subject may train the initial user browsing behavior associated node prediction model based on the browsing behavior causal feature set and the user historical browsing behavior information set to obtain a trained user browsing behavior associated node prediction model. The initial user browsing behavior associated node prediction model may be an untrained DCN (Deep & Cross Network) model or transformer model, or other neural network models.

[0060] In practice, the above execution entity can train the initial user browsing behavior associated node prediction model through the following steps:

[0061] In the first step, the node intervention feature is added to the above-mentioned browsing behavior causal feature set to update the browsing behavior causal feature set. The node intervention feature may be a feature corresponding to the node intervention label.

[0062] The second step is to generate a user historical browsing behavior information training sample set based on the training set corresponding to the above user historical browsing behavior information set and the updated browsing behavior causal feature set. Each user historical browsing behavior information training sample in the above user historical browsing behavior information training sample set includes a response label and each feature information of the corresponding browsing behavior causal feature set, and the feature information corresponding to the node intervention feature is the node intervention label. For each user historical browsing behavior information in the training set, the above execution entity can remove each feature information in the above user historical browsing behavior information except each feature information of the corresponding updated browsing behavior causal feature set to obtain the user historical browsing behavior information training sample.

[0063] The third step is to train the initial user browsing behavior associated node prediction model according to the above user historical browsing behavior information training sample set to obtain a trained user browsing behavior associated node prediction model. For example, first, each user historical browsing behavior information training sample in at least one user historical browsing behavior information training sample in the user historical browsing behavior information training sample set is input into the embedding layer included in the initial user browsing behavior associated node prediction model to obtain each word vector of each user historical browsing behavior information training sample in at least one user historical browsing behavior information training sample corresponding to the above browsing behavior causal feature set. The above embedding layer can be used to generate word vectors (Embedding) of feature information. One browsing behavior causal feature can correspond to one word vector. There is no limitation on the model structure layer after the embedding layer in the initial user browsing behavior associated node prediction model. Second, according to each word vector corresponding to each user historical browsing behavior information training sample in at least one user historical browsing behavior information training sample and the initial user browsing behavior associated node prediction model, a prediction response label corresponding to each user historical browsing behavior information training sample in at least one user historical browsing behavior information training sample is generated. First, input each word vector of each user historical browsing behavior information training sample in the obtained at least one user historical browsing behavior information training sample corresponding to the above-mentioned browsing behavior causal feature set into the feature interaction layer included in the initial user browsing behavior association node prediction model, and obtain the concatenated vector corresponding to each user historical browsing behavior information training sample in the at least one user historical browsing behavior information training sample. Then, input the concatenated word vector corresponding to each user historical browsing behavior information training sample in the obtained at least one user historical browsing behavior information training sample into the prediction layer included in the initial user browsing behavior association node prediction model, and obtain the predicted response label corresponding to each user historical browsing behavior information training sample in the at least one user historical browsing behavior information training sample. Third, compare the predicted response label and the response label corresponding to each user historical browsing behavior information training sample in the at least one user historical browsing behavior information training sample to obtain a comparison result. The above-mentioned execution subject can determine whether the predicted response label and the response label corresponding to each user historical browsing behavior information training sample are the same, so as to generate the accuracy of each predicted response label corresponding to each user historical browsing behavior information training sample as the comparison result. Fourth, in response to determining that the comparison result meets the optimization condition, the initial user browsing behavior associated node prediction model is determined as the trained user browsing behavior associated node prediction model. The above optimization condition may be that the accuracy is greater than or equal to a preset accuracy.

[0064] For another example, the third step may include the following sub-steps:

[0065] The first sub-step is to select a target user historical browsing behavior information training sample from the user historical browsing behavior information training sample set. For example, a user historical browsing behavior information training sample can be randomly selected from the user historical browsing behavior information training sample set as the target user historical browsing behavior information training sample.

[0066] The second sub-step is to group the features of the feature set corresponding to the training samples of the historical browsing behavior information of the target user to generate at least one feature group. The features in the feature set corresponding to the training samples of the historical browsing behavior information of the target user can be randomly grouped into different feature categories to generate at least one feature group. That is, the features between the feature groups are features of different categories. In practice, the feature categories may include: user field description category, page field category, device field category, login behavior description category. The categories of each custom event in the custom event set belong to the user field description category, page field category, device field category, and login behavior description category.

[0067] The third sub-step is to input at least one feature group into the initial feature cross-combination network included in the initial user browsing behavior associated node prediction model to generate feature cross-combination information. The initial feature cross-combination network may be a feature cross-combination model that has not yet been trained. The feature cross-combination network may be a neural network model that cross-combines each feature. For example, the feature cross-combination model may be a multi-layer fully connected layer. The initial user browsing behavior associated node prediction model may be a user browsing behavior associated node prediction model that has not yet been trained. For example, the user browsing behavior associated node prediction model may be a plurality of fully connected modules connected in series. Each fully connected module includes: at least one fully connected layer connected in series. In practice, the feature cross-combination information may be information in the form of a vector.

[0068] The fourth sub-step is to determine whether the initial user browsing behavior associated node prediction model has been trained based on the feature cross-combination information and the sample labels corresponding to the target user's historical browsing behavior information training samples. First, the above-mentioned execution subject can input the feature cross-combination information into the remaining fully connected modules included in the initial user browsing behavior associated node prediction model to generate the initial user browsing behavior associated node prediction information. Then, the loss value between the above-mentioned initial user browsing behavior associated node prediction information and the sample label is determined. Finally, in response to determining that the loss value is less than the preset loss value, it is determined that the initial user browsing behavior associated node prediction model training is completed. In response to determining that the loss value is greater than or equal to the preset loss value, it is determined that the initial user browsing behavior associated node prediction model training is not completed.

[0069] A fifth sub-step, in response to determining that the training is completed, determining the initial user browsing behavior associated node prediction model as the user browsing behavior associated node prediction model.

[0070] Optionally, in response to determining that the training is not completed, the model parameters of the initial user browsing behavior associated node prediction model are adjusted, the adjusted initial user browsing behavior associated node prediction model is used as the initial user browsing behavior associated node prediction model, and the target user historical browsing behavior information training samples are reselected from each user historical browsing behavior information training sample from which the target user historical browsing behavior information training sample is removed, so as to train the initial user browsing behavior associated node prediction model again.

[0071] Therefore, there is no need to use a semi-supervised approach, which reduces label noise and data imbalance caused by the semi-supervised approach. And because there is no need to consider confounding factors, the accuracy of causal inference is improved. Thus, the accuracy of model inference is improved.

[0072] Step 103: input the above-mentioned feature vector into a pre-trained user browsing behavior associated node prediction model to obtain user browsing behavior associated node prediction information.

[0073] In some embodiments, the above-mentioned execution entity can input the above-mentioned feature vector into a pre-trained user browsing behavior associated node prediction model to obtain user browsing behavior associated node prediction information. Among them, the above-mentioned user browsing behavior associated node prediction information can represent the association relationship between the predicted user and each node corresponding to the browsed page. One page corresponds to a node. The node represents the creator of the page. That is, the association relationship can indicate whether there is a communication connection between the user and the creator of the page. For example, the user communicates with the creator of the page through an interface included in the page. The user browsing behavior associated node prediction model can be a pre-trained neural network model that takes the feature vector corresponding to the user as input and the user browsing behavior associated node prediction information as output. The user browsing behavior associated node prediction model can be a trained DCN (Deep & Cross Network) model or transformer model, or it can be other neural network models. The association relationship can refer to a communication relationship.

[0074] Step 104: in response to determining that the node prediction information associated with the user browsing behavior meets a preset condition, push the user information of the user to at least one associated node terminal.

[0075] In some embodiments, the execution entity may push the user information of the user to at least one associated node terminal in response to determining that the user browsing behavior-associated node prediction information meets a preset condition. The preset condition may be: the user browsing behavior-associated node prediction information indicates that the user has a communication connection with the creator of a certain page. The node terminal may refer to the computing device terminal of the page creator. The at least one associated node terminal may refer to a node terminal that has a communication connection with the user. Thus, after pushing the user information to the associated node terminal, the node terminal may push the associated product information or page information to the user based on the user information to improve user stickiness.

[0076] Further references Figure 2 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a browsing behavior information identification device. These browsing behavior information identification device embodiments are Figure 1 Corresponding to the method embodiments shown, the browsing behavior information identification device can be specifically applied to various electronic devices.

[0077] like Figure 2 As shown, the browsing behavior information identification device 200 of some embodiments includes: an acquisition unit 201, a generation unit 202, an input unit 203 and a push unit 204. The acquisition unit 201 is configured to acquire the historical browsing behavior data of the current user within a preset time period, wherein the historical browsing behavior data represents the behavior data of each page browsed by the user; the generation unit 202 is configured to generate a feature vector corresponding to the user according to the historical browsing behavior data; the input unit 203 is configured to input the feature vector into a pre-trained user browsing behavior associated node prediction model to obtain user browsing behavior associated node prediction information, wherein the user browsing behavior associated node prediction information represents the association relationship between the predicted user and each node corresponding to the browsed page; the push unit 204 is configured to push the user information of the user to at least one associated node terminal in response to determining that the user browsing behavior associated node prediction information meets the preset conditions.

[0078] It is understandable that the units recorded in the browsing behavior information identification device 200 and the reference Figure 1 Therefore, the operations, features and beneficial effects described above for the method are also applicable to the browsing behavior information identification device 200 and the units included therein, and will not be described in detail here.

[0079] Reference below Figure 3, which shows a schematic diagram of the structure of an electronic device (e.g., a computing device) 300 suitable for implementing some embodiments of the present disclosure. The electronic devices in some embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (e.g., vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 3 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0080] like Figure 3 As shown, the electronic device 300 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. In the RAM 303, various programs and task data required for the operation of the electronic device 300 are also stored. The processing device 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0081] Typically, the following devices may be connected to the I / O interface 305: input devices 306 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; output devices 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 308 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 309. The communication devices 309 may allow the electronic device 300 to communicate with other devices wirelessly or by wire to exchange task data. Although Figure 3 The electronic device 300 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Figure 3 Each block shown in the figure may represent one device, or may represent multiple devices as required.

[0082] In particular, according to some embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from the network through the communication device 309, or installed from the storage device 308, or installed from the ROM 302. When the computer program is executed by the processing device 301, the above-mentioned functions defined in the method of some embodiments of the present disclosure are executed.

[0083] It should be noted that the computer-readable medium recorded in some embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, 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), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In some embodiments of the present disclosure, a computer-readable signal medium may include a mission data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated mission data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0084] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital task data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0085] The computer-readable medium may be included in the electronic device; or it may exist independently without being installed in the electronic device. The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: obtains the historical browsing behavior data of the current user within a preset time period, wherein the historical browsing behavior data represents the behavior data of each page browsed by the user; generates a feature vector corresponding to the user based on the historical browsing behavior data; inputs the feature vector into a pre-trained user browsing behavior associated node prediction model to obtain user browsing behavior associated node prediction information, wherein the user browsing behavior associated node prediction information represents the association relationship between the predicted user and each node corresponding to the browsed page; in response to determining that the user browsing behavior associated node prediction information meets the preset conditions, pushes the user information of the user to at least one associated node terminal.

[0086] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including product-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving 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., via the Internet using an Internet service provider).

[0087] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0088] The units described in some embodiments of the present disclosure may be implemented by software or by hardware. The described units may also be set in a processor, for example, may be described as: a processor comprising: an acquisition unit, a generation unit, an input unit and a push unit. Among them, the names of these units do not constitute a limitation on the unit itself under certain circumstances. For example, the push unit may also be described as "a unit that pushes the user information of the above-mentioned user to at least one associated node terminal in response to determining that the node prediction information associated with the above-mentioned user browsing behavior meets the preset conditions."

[0089] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0090] The above descriptions are only some preferred embodiments of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, the above-mentioned features are replaced with the technical features with similar functions disclosed in the embodiments of the present disclosure (but not limited to) and the technical solutions formed.

Claims

1. A method for identifying browsing behavior information, comprising: Obtaining historical browsing behavior data of the current user within a preset time period, wherein the historical browsing behavior data represents the behavior data of each page browsed by the user; Generating a feature vector corresponding to the user according to the historical browsing behavior data; Acquire a user historical browsing behavior information set, wherein the user historical browsing behavior information in the user historical browsing behavior information set includes a response tag, a node intervention tag, and a feature information set; According to the user historical browsing behavior information set, a browsing response label classification network and a node intervention label classification network are trained, wherein the browsing response label classification network and the node intervention label classification network both correspond to feature sequences; generating a browsing behavior causal feature set according to a feature sequence corresponding to the browsing response label classification network and a feature sequence corresponding to the node intervention label classification network; According to the browsing behavior causal feature set and the user historical browsing behavior information set, the initial user browsing behavior associated node prediction model is trained to obtain a trained user browsing behavior associated node prediction model; Inputting the feature vector into a pre-trained user browsing behavior associated node prediction model to obtain user browsing behavior associated node prediction information, wherein the user browsing behavior associated node prediction information represents the association relationship between the predicted user and each node corresponding to the browsed page; In response to determining that the user browsing behavior-associated node prediction information meets a preset condition, pushing the user information of the user to at least one associated node terminal; The step of generating a feature vector corresponding to the user according to the historical browsing behavior data includes: The custom event set is determined by the following formula: S={(A1,c1,t1),(A2,c2,t2)…,(A n ,c n ,t n )}, Among them, S represents a custom event set, A1, A2, A n is a custom event set, c1, c2, c n Indicates the access time to the custom event, t1, t2, t n Indicates the residence time of a custom event; Determine the total access time for a custom event set using the following formula: Wherein, C is the total access time, c i is the access time to the custom event, n is the number of custom events in the custom event set; The total dwell time of a custom event set is determined by the following formula: Wherein, T is the total residence time, t i The dwell time for custom events; The proportion of each access time in the total access time is determined by the following formula: in, Indicates the proportion of the i-th access time in the total access time; Determine the contribution of each dwell time to the total dwell time using the following formula: in, is the proportion of the i-th residence time in the total residence time; The user's interest in each custom event is determined using the following formula: Among them, ω i is the i-th interest degree; The custom event set, the total access time, the total dwell time, the proportion of each access time in the total access time, the proportion of each dwell time in the total dwell time, and the user's interest in each custom event are combined into a feature vector corresponding to the user; The step of generating a browsing behavior causal feature set according to the feature sequence corresponding to the browsing response label classification network and the feature sequence corresponding to the node intervention label classification network includes: Determine each feature that satisfies a preset feature condition in the feature sequence corresponding to the browsing response label classification network as a first feature group; Determine each feature that satisfies a preset feature condition in the feature sequence corresponding to the node intervention classification network as a second feature group; determining the intersection of the first feature group and the second feature group as a browsing behavior causal feature set; The step of training the initial user browsing behavior associated node prediction model according to the browsing behavior causal feature set and the user historical browsing behavior information set to obtain the trained user browsing behavior associated node prediction model includes: Adding the node intervention feature to the browsing behavior causal feature set to update the browsing behavior causal feature set; Generate a user historical browsing behavior information training sample set according to the training set corresponding to the user historical browsing behavior information set and the updated browsing behavior causal feature set, wherein each user historical browsing behavior information training sample in the user historical browsing behavior information training sample set includes a response label and each feature information of the corresponding browsing behavior causal feature set, and the feature information corresponding to the node intervention feature is the node intervention label; Training the initial user browsing behavior associated node prediction model based on the user historical browsing behavior information training sample set to obtain a trained user browsing behavior associated node prediction model; The training of the initial user browsing behavior associated node prediction model based on the user historical browsing behavior information training sample set to obtain the trained user browsing behavior associated node prediction model includes: Selecting a target user's historical browsing behavior information training sample from the user's historical browsing behavior information training sample set; Performing feature grouping on the feature set corresponding to the training sample of the target user's historical browsing behavior information to generate at least one feature group; Inputting at least one feature group into an initial feature cross-combination network included in the initial user browsing behavior associated node prediction model to generate feature cross-combination information; Determine whether the initial user browsing behavior associated node prediction model has been trained based on the sample labels corresponding to the feature cross-combination information and the target user's historical browsing behavior information training samples; In response to determining that the training is completed, the initial user browsing behavior associated node prediction model is determined as the user browsing behavior associated node prediction model.

2. A browsing behavior information identification device, applied to the method according to claim 1, comprising: An acquisition unit is configured to acquire historical browsing behavior data of the current user within a preset time period, wherein the historical browsing behavior data represents behavior data of each page browsed by the user; A generating unit, configured to generate a feature vector corresponding to the user according to the historical browsing behavior data; An input unit is configured to input the feature vector into a pre-trained user browsing behavior associated node prediction model to obtain user browsing behavior associated node prediction information, wherein the user browsing behavior associated node prediction information represents the predicted association relationship between the user and each node corresponding to the browsed page; The push unit is configured to push the user information of the user to at least one associated node terminal in response to determining that the node prediction information associated with the user browsing behavior meets a preset condition.

3. An electronic device, comprising: one or more processors; a storage device having one or more programs stored thereon; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as claimed in claim 1.

4. A computer readable medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method according to claim 1 is implemented.

Images