Information push method, device, electronic device, and computer-readable medium

Abstract

The embodiments of the present disclosure disclose an information push method, device, electronic device, and computer-readable medium. A specific implementation of the method includes: splitting the feedback user feature vector set corresponding to the feedback target into each first-level feedback user feature vector subset based on time granularity information; for each first-level feedback user feature vector subset, clustering each first-level feedback user feature vector to obtain each second-level feedback user feature vector subset; performing optimization processing on each obtained second-level feedback user feature vector subset to obtain a seed user feature vector set; generating similarity information in response to receiving the target user feature vector; and pushing preset channel information to the user terminal corresponding to the target user feature vector in response to the similarity information meeting a preset similarity condition. This implementation improves the accuracy and recall rate of the selected seed users, enhances the association effect of amplifying similar groups, and saves computing resources.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of computer technology, and more particularly to an information push method, apparatus, electronic device, and computer-readable medium. Background Art

[0002] Crowd lookalike technology is a technique that uses seed users to find more similar groups of people with potential connections. Currently, seed users are usually selected manually or by using real-time feedback users as seed users.

[0003] However, when using the above method to select seed users, the following technical problems often arise: when there are many user portrait tags, the accuracy and recall rate of seed users selected manually are low; if the feedback users are directly used as seed users, the characteristics of the seed users are relatively scattered, and the association effect of amplifying similar groups is poor. At the same time, when there are many feedback users, searching for similar groups results in a waste of computing resources. Summary of the Invention

[0004] The content of this disclosure is used to briefly introduce concepts that will be described in detail in the detailed description section below. The content of this disclosure is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0005] Some embodiments of the present disclosure provide information push methods, devices, electronic devices, and computer-readable media to solve one or more of the technical problems mentioned in the above background technology section.

[0006] In a first aspect, some embodiments of the present disclosure provide an information push method, the method comprising: splitting a feedback user feature vector set corresponding to a feedback target into first-level feedback user feature vector subsets according to preset time granularity information, wherein each first-level feedback user feature vector subset corresponds to a time period; for each first-level feedback user feature vector subset in the first-level feedback user feature vector subsets, performing clustering processing on each first-level feedback user feature vector in the first-level feedback user feature vector subset to obtain second-level feedback user feature vector subsets; performing second-level feedback user feature vector subset optimization processing on each obtained second-level feedback user feature vector subset to obtain a selected second-level feedback user feature vector subset as a seed user feature vector set; in response to receiving a target user feature vector, generating similarity information based on the seed user feature vector set and the target user feature vector; and in response to the similarity information satisfying a preset similarity condition, pushing preset channel information corresponding to the feedback target to a user terminal corresponding to the target user feature vector.

[0007] Optionally, the method further includes: in response to the similarity information satisfying a preset similarity condition, adding the target user feature vector to the seed user feature vector set.

[0008] Optionally, the generating of similarity information includes: performing averaging processing on each seed user feature vector in the seed user feature vector set to obtain a mean seed user feature vector; and determining the similarity between the target user feature vector and the mean seed user feature vector as the similarity information.

[0009] Optionally, before splitting the feedback user feature vector set corresponding to the feedback target into first-level feedback user feature vector subsets according to the preset time granularity information, the method also includes: in response to receiving the target burial point log data corresponding to the above feedback target, parsing and processing the above target burial point log data to obtain the feedback user feature vector set.

[0010] Optionally, the above-mentioned parsing and processing of the above-mentioned target burial point log data includes: performing field parsing processing on the above-mentioned target burial point log data to obtain a parsed user feature vector set; in response to the existence of the same parsed user feature vector in the above-mentioned parsed user feature vector set, deduplicating the above-mentioned parsed user feature vector set to obtain a deduplicated parsed user feature vector set as a feedback user feature vector set; in response to the differences between the individual parsed user feature vectors in the above-mentioned parsed user feature vector set, determining the above-mentioned parsed user feature vector set as a feedback user feature vector set.

[0011] Optionally, the method further includes: in response to the feedback amount corresponding to the seed user feature vector in the above-mentioned seed user feature vector set detected within a preset time period meeting a preset feedback abnormality condition, deleting the seed user feature vector corresponding to the above-mentioned feedback amount from the above-mentioned seed user feature vector set, wherein the above-mentioned feedback amount corresponds to the above-mentioned feedback target.

[0012] In a second aspect, some embodiments of the present disclosure provide an information push device, comprising: a splitting unit configured to split a feedback user feature vector set corresponding to a feedback target into first-level feedback user feature vector subsets according to preset time granularity information, wherein each first-level feedback user feature vector subset corresponds to a time period; a clustering unit configured to perform clustering processing on each first-level feedback user feature vector subset in each first-level feedback user feature vector subset to obtain second-level feedback user feature vector subsets; a selecting unit configured to perform second-level feedback user feature vector subset selecting processing on each obtained second-level feedback user feature vector subset to obtain a selected second-level feedback user feature vector subset as a seed user feature vector set; a generating unit configured to generate similarity information based on the seed user feature vector set and the target user feature vector in response to receiving a target user feature vector; and a pushing unit configured to push preset channel information corresponding to the feedback target to a user terminal corresponding to the target user feature vector in response to the similarity information satisfying a preset similarity condition.

[0013] Optionally, the apparatus further comprises: an adding unit configured to add the target user feature vector to the seed user feature vector set in response to the similarity information satisfying a preset similarity condition.

[0014] Optionally, the generating unit is further configured to: perform averaging processing on each seed user feature vector in the seed user feature vector set to obtain a mean seed user feature vector; and determine the similarity between the target user feature vector and the mean seed user feature vector as similarity information.

[0015] Optionally, before the splitting unit, the device also includes: a parsing unit, which is configured to parse and process the above-mentioned target burial point log data in response to receiving the target burial point log data corresponding to the above-mentioned feedback target to obtain a feedback user feature vector set.

[0016] Optionally, the parsing unit is further configured to: perform field parsing on the above-mentioned target embedded point log data to obtain a parsed user feature vector set; in response to the existence of identical parsed user feature vectors in the above-mentioned parsed user feature vector set, perform deduplication processing on the above-mentioned parsed user feature vector set to obtain a deduplicated parsed user feature vector set as a feedback user feature vector set; in response to the differences between the individual parsed user feature vectors in the above-mentioned parsed user feature vector set, determine the above-mentioned parsed user feature vector set as a feedback user feature vector set.

[0017] Optionally, the device further includes: a deletion unit configured to delete the seed user feature vector corresponding to the feedback amount from the seed user feature vector set in response to the feedback amount corresponding to the seed user feature vector in the seed user feature vector set detected within a preset time period meeting a preset feedback abnormality condition, wherein the feedback amount corresponds to the feedback target.

[0018] 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 one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0019] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method described in any implementation of the first aspect is implemented.

[0020] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the information push method of some embodiments of the present disclosure, the accuracy and recall rate of the selected seed users are improved, the correlation effect of similar group amplification is enhanced, and computing power resources are saved. Specifically, the reasons for the low accuracy and recall rate of the selected seed users, the poor correlation effect of similar group amplification, and the waste of computing power resources are: when there are many user portrait tags, the accuracy and recall rate of the seed users selected manually are low; when the feedback users are directly used as seed users, the characteristics of the seed users are relatively scattered, and the correlation effect of similar group amplification is poor. At the same time, when there are many feedback users, similar group search is performed, resulting in a waste of computing power resources. Based on this, the information recommendation method of some embodiments of the present disclosure, first, according to the preset time granularity information, the feedback user feature vector set corresponding to the feedback target is split into each first-level feedback user feature vector subset, wherein each first-level feedback user feature vector subset corresponds to a time period. Thus, the real-time feedback user feature vector set can be split into multiple first-level feedback user feature vector subsets from a horizontal (time) perspective. Then, for each of the first-level feedback user feature vector subsets in the first-level feedback user feature vector subsets, clustering processing is performed on each of the first-level feedback user feature vector subsets to obtain second-level feedback user feature vector subsets. Thus, each first-level feedback user feature vector subset can be further split into second-level feedback user feature vector subsets from a vertical (clustering category) perspective. Second-level feedback user feature vector subsets are then optimized for each of the obtained second-level feedback user feature vector subsets to obtain a selected second-level feedback user feature vector subset as a seed user feature vector set. Thus, a second-level user feedback feature vector subset can be optimized from each of the second-level feedback user feature vector subsets obtained through the horizontal and vertical splitting to serve as a seed user feature vector set. Next, in response to receiving a target user feature vector, similarity information is generated based on the seed user feature vector set and the target user feature vector. Finally, in response to the similarity information satisfying a preset similarity condition, preset channel information corresponding to the feedback target is pushed to the user terminal corresponding to the target user feature vector. This allows the user terminal to display the preset channel information in the resource section of the application page. Because the seed user feature vector set is selected from the subset of secondary feedback user feature vectors obtained through horizontal and vertical splitting, the seed user feature vector set achieves clustering of features. This improves the correlation effect of amplifying similar groups. Furthermore, because the entire feedback user feature vector set is not used as the seed user feature vector set, computing resources can be saved when searching for similar groups. Furthermore, the automatic selection of the seed user feature vector set improves the accuracy and recall rate of the selected seed users. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and other features, advantages, and aspects of the various embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the 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.

[0022] Figure 1 is a schematic diagram of an application scenario of the information push method according to some embodiments of the present disclosure;

[0023] Figure 2 is a flowchart of some embodiments of the information push method according to the present disclosure;

[0024] Figure 3 is a flowchart of other embodiments of the information push method according to the present disclosure;

[0025] Figure 4 is a flowchart of some further embodiments of the information push method according to the present disclosure;

[0026] Figure 5 is a schematic structural diagram of some embodiments of the information push device according to the present disclosure;

[0027] Figure 6 It is a structural diagram of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0028] 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. 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 for illustrative purposes only and are not intended to limit the scope of protection of the present disclosure.

[0029] 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 may be combined with each other.

[0030] It should be noted that the concepts of "first" and "second" mentioned in this 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.

[0031] It should be noted that the modifications of "one" and "multiple" 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, they should be understood as "one or more".

[0032] 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.

[0033] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0034] Figure 1 It is a schematic diagram of an application scenario of the information recommendation method according to some embodiments of the present disclosure.

[0035] exist Figure 1 In the application scenario, first, the computing device 101 can split the feedback user feature vector set 104 corresponding to the feedback target 103 into each first-level feedback user feature vector subset 105 according to the preset time granularity information 102. Each first-level feedback user feature vector subset corresponds to a time period. Then, for each first-level feedback user feature vector subset in the above-mentioned first-level feedback user feature vector subsets 105, the computing device 101 can perform clustering processing on each first-level feedback user feature vector in the above-mentioned first-level feedback user feature vector subset to obtain each second-level feedback user feature vector subset. Thereafter, the computing device 101 can perform second-level feedback user feature vector subset optimization processing on each obtained second-level feedback user feature vector subset 106, and obtain the second-level feedback user feature vector subset after optimization processing as the seed user feature vector set 107. Secondly, in response to receiving the target user feature vector 108, the computing device 101 can generate similarity information 109 based on the above-mentioned seed user feature vector set 107 and the above-mentioned target user feature vector 108. Finally, in response to the similarity information 109 satisfying a preset similarity condition, the computing device 101 may push the preset channel information 110 corresponding to the feedback target 103 to the user terminal 111 corresponding to the target user feature vector 108 .

[0036] It should be noted that the computing device 101 can be either hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or a single terminal device. When the computing device is software, it can be installed in the hardware devices listed above. It can be implemented as multiple software or software modules for providing distributed services, or as a single software or software module. No specific limitations are given here.

[0037] It should be understood that Figure 1 The number of computing devices in the embodiment is merely illustrative. Any number of computing devices may be provided according to implementation requirements.

[0038] Continue to refer Figure 2 , shows a process 200 of some embodiments of the information push method according to the present disclosure. The information push method includes the following steps:

[0039] Step 201 : Split the feedback user feature vector set corresponding to the feedback target into first-level feedback user feature vector subsets according to preset time granularity information.

[0040] In some embodiments, the execution subject of the information push method (eg Figure 1 The computing device 101 shown) can split the feedback user feature vector set corresponding to the feedback target into each first-level feedback user feature vector subset according to the preset time granularity information. The above-mentioned time granularity information can be pre-set time granularity related information for splitting the feedback user feature vector set. For example, the above-mentioned time granularity information can be "10 minutes". The above-mentioned feedback target can be a target for the behavior of the corresponding user in the front-end application page. For example, the feedback target can be "increasing the click-through exposure rate of the application page". The above-mentioned feedback user feature vector set can be a collection of feature vectors of each user who clicks to enter the above-mentioned application page. The feedback user feature vectors in the above-mentioned feedback user feature vector set can include but are not limited to: user attribute information, user behavior information. The above-mentioned user attribute information can be information related to the basic attributes of the user, and can be but are not limited to: age, gender, city. The above-mentioned user behavior information can be information related to the user's behavior on the application page, and can be but are not limited to: the number of visits, active days, and visit duration within a preset historical time period. In practice, the execution entity may split the feedback user feature vector set corresponding to the feedback target into first-level feedback user feature vector subsets according to the time granularity represented by the time granularity information, such that each first-level feedback user feature vector subset corresponds to a time period, and the time intervals of the time periods are the same. For example, if the time granularity information is "10 minutes," the time intervals of the time periods are 10 minutes.

[0041] Optionally, the above-mentioned execution subject may parse and process the above-mentioned target buried point log data in response to receiving the target buried point log data corresponding to the above-mentioned feedback target, and obtain a feedback user feature vector set. Among them, the above-mentioned target buried point log data may be log data received based on the buried point of the target user behavior corresponding to the application page. The above-mentioned application page and the above-mentioned target user behavior correspond to the above-mentioned feedback target. For example, the above-mentioned feedback target is "to improve the click exposure rate of the application page". Then the above-mentioned target user behavior is "the click behavior of the user clicking to enter the above-mentioned application page". In practice, for each user identifier, the above-mentioned execution subject may extract the various feature values ​​corresponding to the various fields in the above-mentioned field set from the above-mentioned target buried point log data according to a preset field set as the feedback user feature vector corresponding to the above-mentioned user identifier. In this way, the feedback user feature vector set can be automatically parsed through the target buried point log data.

[0042] Optionally, first, the above-mentioned execution subject may perform field parsing processing on the above-mentioned target buried point log data to obtain a parsed user feature vector set. In practice, for each user identifier, the above-mentioned execution subject may extract, from the above-mentioned target buried point log data, the respective feature values ​​corresponding to the respective fields in the above-mentioned field set according to the preset field set as the parsed user feature vector corresponding to the above-mentioned user identifier. Then, in response to the existence of identical parsed user feature vectors in the above-mentioned parsed user feature vector set, the above-mentioned execution subject may perform deduplication processing on the above-mentioned parsed user feature vector set to obtain the deduplicated parsed user feature vector set as the feedback user feature vector set. Thereafter, in response to the differences between the respective parsed user feature vectors in the above-mentioned parsed user feature vector set, the above-mentioned execution subject may determine the above-mentioned parsed user feature vector set as the feedback user feature vector set. In this way, it is possible to ensure that identical feedback user feature vectors do not exist in the feedback user feature vector set.

[0043] Step 202 : For each first-level feedback user feature vector subset in each first-level feedback user feature vector subset, clustering processing is performed on each first-level feedback user feature vector in the first-level feedback user feature vector subset to obtain each second-level feedback user feature vector subset.

[0044] In some embodiments, the execution entity may perform clustering on each of the first-level feedback user feature vector subsets in the first-level feedback user feature vector subsets to obtain each second-level feedback user feature vector subset. In practice, the execution entity may employ a kmeans clustering algorithm to perform clustering on each of the first-level feedback user feature vector subsets.

[0045] Step 203 : performing secondary feedback user feature vector subset optimization processing on each obtained secondary feedback user feature vector subset, and obtaining the secondary feedback user feature vector subset after optimization processing as a seed user feature vector set.

[0046] In some embodiments, the execution entity may perform secondary feedback user feature vector subset optimization on each obtained secondary feedback user feature vector subset, obtaining the optimized secondary feedback user feature vector subset as the seed user feature vector set. In practice, the execution entity may select the optimal secondary feedback user feature vector subset from each obtained secondary feedback user feature vector subset using a Bandit algorithm based on UCB (Upper Confidence Bounds) as the seed user feature vector set.

[0047] Step 204 : In response to receiving the target user feature vector, similarity information is generated based on the seed user feature vector set and the target user feature vector.

[0048] In some embodiments, the execution subject of the information push method (eg Figure 1 The computing device 101 shown in the figure can generate similarity information based on the above-mentioned seed user feature vector set and the above-mentioned target user feature vector in response to receiving the target user feature vector. The above-mentioned target user feature vector can be a feature vector of a user who visits the application page of the above-mentioned feedback target in real time. It can be understood that the dimension of the above-mentioned target user feature vector is consistent with that of the seed user feature vector in the above-mentioned seed user feature vector set. In practice, the above-mentioned execution entity can use the cosine similarity formula to generate the similarity between each seed user feature vector in the above-mentioned seed user feature vector set and the above-mentioned target user feature vector, and obtain the similarity set as the similarity information.

[0049] In some optional implementations of some embodiments, first, the execution entity may perform averaging on each seed user feature vector in the seed user feature vector set to obtain a mean seed user feature vector. In practice, for each dimension of the seed user feature vector in the seed user feature vector set, the execution entity may determine the mean of the eigenvalues ​​of each seed user feature vector in the seed user feature vector set corresponding to the dimension as the eigenvalue of the mean seed user feature vector corresponding to the dimension, thereby obtaining each eigenvalue of the mean seed user feature vector corresponding to each dimension. Then, the similarity between the target user feature vector and the mean seed user feature vector may be determined as similarity information. In practice, the execution entity may use the cosine similarity formula to generate the similarity between the target user feature vector and the mean seed user feature vector as similarity information. The execution entity may also use a multi-layer perceptron (MLP) to generate the similarity between the target user feature vector and the mean seed user feature vector as similarity information.

[0050] Step 205 : In response to the similarity information satisfying a preset similarity condition, the preset channel information corresponding to the feedback target is pushed to the user terminal corresponding to the target user feature vector.

[0051] In some embodiments, in response to the similarity information satisfying a preset similarity condition, the execution entity may push the preset channel information corresponding to the feedback target to the user terminal corresponding to the target user feature vector. The preset channel information may be related information for displaying information that can be linked to the application page corresponding to the feedback target, and may include the application page URL, channel rendering information, and resource position information. The channel rendering information may be information for display on the application page, and may include a channel picture. The resource position information may be location information for rendering the channel rendering information on the application page. The user terminal may be a terminal device of a user who visits the application page in real time. The preset similarity condition may be "the median of the similarity included in the similarity information is greater than or equal to a preset threshold value." Thus, the user terminal may display the preset channel information in the resource position of the application page.

[0052] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the information push method of some embodiments of the present disclosure, the accuracy and recall rate of the selected seed users are improved, the correlation effect of similar group amplification is enhanced, and computing power resources are saved. Specifically, the reasons for the low accuracy and recall rate of the selected seed users, the poor correlation effect of similar group amplification, and the waste of computing power resources are: when there are many user portrait tags, the accuracy and recall rate of the seed users selected manually are low; when the feedback users are directly used as seed users, the characteristics of the seed users are relatively scattered, and the correlation effect of similar group amplification is poor. At the same time, when there are many feedback users, similar group search is performed, resulting in a waste of computing power resources. Based on this, the information recommendation method of some embodiments of the present disclosure, first, according to the preset time granularity information, the feedback user feature vector set corresponding to the feedback target is split into each first-level feedback user feature vector subset, wherein each first-level feedback user feature vector subset corresponds to a time period. Thus, the real-time feedback user feature vector set can be split into multiple first-level feedback user feature vector subsets from a horizontal (time) perspective. Then, for each of the first-level feedback user feature vector subsets in the first-level feedback user feature vector subsets, clustering processing is performed on each of the first-level feedback user feature vector subsets to obtain second-level feedback user feature vector subsets. Thus, each first-level feedback user feature vector subset can be further split into second-level feedback user feature vector subsets from a vertical (clustering category) perspective. Second-level feedback user feature vector subsets are then optimized for each of the obtained second-level feedback user feature vector subsets to obtain a selected second-level feedback user feature vector subset as a seed user feature vector set. Thus, a second-level user feedback feature vector subset can be optimized from each of the second-level feedback user feature vector subsets obtained through the horizontal and vertical splitting to serve as a seed user feature vector set. Next, in response to receiving a target user feature vector, similarity information is generated based on the seed user feature vector set and the target user feature vector. Finally, in response to the similarity information satisfying a preset similarity condition, preset channel information corresponding to the feedback target is pushed to the user terminal corresponding to the target user feature vector. This allows the user terminal to display the preset channel information in the resource section of the application page. Because the seed user feature vector set is selected from the subset of secondary feedback user feature vectors obtained through horizontal and vertical splitting, the seed user feature vector set achieves clustering of features. This improves the correlation effect of amplifying similar groups. Furthermore, because the entire feedback user feature vector set is not used as the seed user feature vector set, computing resources can be saved when searching for similar groups. Furthermore, the automatic selection of the seed user feature vector set improves the accuracy and recall rate of the selected seed users.

[0053] Further references Figure 3 , which shows a process 300 of another embodiment of the information push method. The process 300 of the information push method includes the following steps:

[0054] Step 301 : Split the feedback user feature vector set corresponding to the feedback target into first-level feedback user feature vector subsets according to preset time granularity information.

[0055] Step 302 : For each first-level feedback user feature vector subset in each first-level feedback user feature vector subset, clustering processing is performed on each first-level feedback user feature vector in the first-level feedback user feature vector subset to obtain each second-level feedback user feature vector subset.

[0056] Step 303 : performing secondary feedback user feature vector subset optimization processing on each obtained secondary feedback user feature vector subset, and obtaining the secondary feedback user feature vector subset after optimization processing as a seed user feature vector set.

[0057] Step 304 : In response to receiving the target user feature vector, similarity information is generated based on the seed user feature vector set and the target user feature vector.

[0058] Step 305 : In response to the similarity information satisfying a preset similarity condition, the preset channel information corresponding to the feedback target is pushed to the user terminal corresponding to the target user feature vector.

[0059] In some embodiments, the specific implementation of steps 301-305 and the resulting technical effects can be referred to Figure 2 The corresponding steps 201-205 in the embodiments are not described in detail here.

[0060] Step 306 : In response to the feedback amount corresponding to the seed user feature vector in the seed user feature vector set detected within the preset time period meeting the preset feedback abnormality condition, the seed user feature vector corresponding to the feedback amount is deleted from the seed user feature vector set.

[0061] In some embodiments, the execution subject of the information push method (eg Figure 1The computing device 101 shown) can delete the seed user feature vector corresponding to the above-mentioned feedback amount from the above-mentioned seed user feature vector set in response to the feedback amount corresponding to the seed user feature vector in the above-mentioned seed user feature vector set detected within a preset time period meeting the preset feedback abnormality condition. The above-mentioned feedback amount corresponds to the above-mentioned feedback target. The above-mentioned feedback amount can be a statistical quantity related to the above-mentioned feedback target. For example, the above-mentioned feedback target can be "increasing the click-through exposure rate of the application page". Then the above-mentioned feedback amount can be the number of times the user clicks to enter the above-mentioned application page within the above-mentioned preset time period. The above-mentioned preset feedback abnormality condition can be "the above-mentioned feedback amount is less than or equal to the preset feedback amount". Here, there is no limitation on the specific setting of the preset feedback amount.

[0062] from Figure 3 It can be seen that Figure 2 Compared with the description of some corresponding embodiments, Figure 3 Process 300 of the information push method in some corresponding embodiments embodies the step of deleting seed user feature vectors whose feedback volume meets the preset feedback anomaly condition. Thus, the schemes described in these embodiments can dynamically adjust the set of seed user feature vectors, preventing the number of seed user feature vectors in the set from increasing, further conserving computing resources. Furthermore, deleting seed user feature vectors whose feedback volume meets the preset feedback anomaly condition further improves the accuracy of selecting seed users, thereby enhancing the association effect of amplifying similar groups.

[0063] Further references Figure 4 , which shows a process 400 of some further embodiments of the information push method. The process 400 of the information push method includes the following steps:

[0064] Step 401 : Split the feedback user feature vector set corresponding to the feedback target into first-level feedback user feature vector subsets according to preset time granularity information.

[0065] Step 402 : For each first-level feedback user feature vector subset in each first-level feedback user feature vector subset, clustering processing is performed on each first-level feedback user feature vector in the first-level feedback user feature vector subset to obtain each second-level feedback user feature vector subset.

[0066] Step 403 : performing secondary feedback user feature vector subset optimization processing on each obtained secondary feedback user feature vector subset, and obtaining the secondary feedback user feature vector subset after optimization processing as a seed user feature vector set.

[0067] Step 404 : In response to receiving the target user feature vector, similarity information is generated based on the seed user feature vector set and the target user feature vector.

[0068] Step 405 : In response to the similarity information satisfying a preset similarity condition, the preset channel information corresponding to the feedback target is pushed to the user terminal corresponding to the target user feature vector.

[0069] In some embodiments, the specific implementation of steps 401-405 and the resulting technical effects can be referred to Figure 2 The corresponding steps 201-205 in the embodiments are not described in detail here.

[0070] Step 406 : In response to the similarity information satisfying a preset similarity condition, the target user feature vector is added to the seed user feature vector set.

[0071] In some embodiments, the execution entity may add the target user feature vector to the set of seed user feature vectors in response to the similarity information satisfying the preset similarity condition. The preset similarity condition may be "the median of the similarities included in the similarity information is greater than or equal to a preset threshold." The specific setting of the preset threshold is not limited. Thus, the set of seed user feature vectors can be expanded based on the user feature vectors of users visiting in real time.

[0072] It should be noted that steps 401-403 only need to be performed once. When the target user feature vector is received again, the seed user feature vector set obtained in step 406 can be directly used to perform steps 404-405 again.

[0073] from Figure 4 It can be seen that Figure 2 Compared with the description of some corresponding embodiments, Figure 4 The process 400 of the information push method in some corresponding embodiments embodies the step of expanding the seed user feature vector set. Therefore, the solutions described in these embodiments can expand the seed user feature vector set based on the user feature vectors of users who visit in real time.

[0074] Further references Figure 5 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of an information push device. These device embodiments are similar to Figure 2 Corresponding to the method embodiments shown, the device can be specifically applied to various electronic devices.

[0075] like Figure 5As shown, the information pushing device 500 of some embodiments includes: a splitting unit 501, a clustering unit 502, a optimization unit 503, a generation unit 504 and a pushing unit 505. The splitting unit 501 is configured to split the feedback user feature vector set corresponding to the feedback target into first-level feedback user feature vector subsets according to preset time granularity information, where each first-level feedback user feature vector subset corresponds to a time period. The clustering unit 502 is configured to perform clustering processing on each first-level feedback user feature vector subset in the first-level feedback user feature vector subsets to obtain second-level feedback user feature vector subsets. The selecting unit 503 is configured to perform second-level feedback user feature vector subset selection processing on the obtained second-level feedback user feature vector subsets to obtain the selected second-level feedback user feature vector subsets as the seed user feature vector set. The generating unit 504 is configured to generate similarity information based on the seed user feature vector set and the target user feature vector in response to receiving the target user feature vector. The pushing unit 505 is configured to push the preset channel information corresponding to the feedback target to the user terminal corresponding to the target user feature vector in response to the similarity information satisfying a preset similarity condition.

[0076] Optionally, the information pushing device 500 may further include: an adding unit (not shown in the figure), configured to add the target user feature vector to the seed user feature vector set in response to the similarity information meeting a preset similarity condition.

[0077] Optionally, the generating unit 504 may be further configured to: perform averaging processing on each seed user feature vector in the seed user feature vector set to obtain a mean seed user feature vector; and determine the similarity between the target user feature vector and the mean seed user feature vector as similarity information.

[0078] Optionally, before the splitting unit 501, the information push device 500 may also include: a parsing unit (not shown in the figure), which is configured to parse and process the above-mentioned target burial point log data in response to receiving the target burial point log data corresponding to the above-mentioned feedback target to obtain a feedback user feature vector set.

[0079] Optionally, the parsing unit can be further configured to: perform field parsing on the above-mentioned target embedded point log data to obtain a parsed user feature vector set; in response to the existence of identical parsed user feature vectors in the above-mentioned parsed user feature vector set, perform deduplication processing on the above-mentioned parsed user feature vector set to obtain a deduplicated parsed user feature vector set as a feedback user feature vector set; in response to the differences between the individual parsed user feature vectors in the above-mentioned parsed user feature vector set, determine the above-mentioned parsed user feature vector set as a feedback user feature vector set.

[0080] Optionally, the information push device 500 may further include: a deletion unit (not shown in the figure), configured to delete the seed user feature vector corresponding to the feedback amount from the above-mentioned seed user feature vector set in response to the feedback amount corresponding to the seed user feature vector in the above-mentioned seed user feature vector set detected within a preset time period meeting a preset feedback abnormality condition, wherein the above-mentioned feedback amount corresponds to the above-mentioned feedback target.

[0081] It is understood that the units described in the device 500 are similar to those in the reference Figure 2 Therefore, the operations, features and beneficial effects described above for the method are also applicable to the device 500 and the units included therein, and will not be repeated here.

[0082] Reference below Figure 6 , which shows an electronic device (eg, Figure 1 Schematic diagram of the structure of the computing device 101)600. Figure 6 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.

[0083] like Figure 6 As shown, the electronic device 600 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. Various programs and data required for the operation of the electronic device 600 are also stored in the RAM 603. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0084] Typically, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 6 The electronic device 600 is shown with various devices, but it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed instead. Figure 6 Each block shown in the figure may represent one device, or may represent multiple devices as needed.

[0085] In particular, according to some embodiments of the present disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of the present disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program comprising program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from a network via the communication device 609, or installed from the storage device 608, or installed from the ROM 602. When the computer program is executed by the processing device 601, the above-mentioned functions defined in the method of some embodiments of the present disclosure are performed.

[0086] It should be noted that the computer-readable medium described 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 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, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, device, or device. In some embodiments of the present disclosure, the computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport 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 thereof.

[0087] In some embodiments, the client and server can communicate using any currently known or future developed network protocol, such as HTTP (HyperText Transfer Protocol), and can be interconnected with any form or medium of digital 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.

[0088] The computer-readable medium may be included in the electronic device, or may exist independently and not be incorporated into the electronic device. The computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: split the feedback user feature vector set corresponding to the feedback target into first-level feedback user feature vector subsets according to preset time granularity information, wherein each first-level feedback user feature vector subset corresponds to a time period; perform clustering processing on each first-level feedback user feature vector subset in the first-level feedback user feature vector subset to obtain second-level feedback user feature vector subsets; perform second-level feedback user feature vector subset optimization processing on each obtained second-level feedback user feature vector subset to obtain the optimized second-level feedback user feature vector subset as a seed user feature vector set; generate similarity information based on the seed user feature vector set and the target user feature vector in response to receiving the target user feature vector; and push the preset channel information corresponding to the feedback target to the user terminal corresponding to the target user feature vector in response to the similarity information satisfying a preset similarity condition.

[0089] 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 object-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 stand-alone 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 through 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., through the Internet using an Internet service provider).

[0090] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the module, program segment, or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order than that marked in the accompanying drawings. For example, two 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 box in the block diagram and / or flowchart, and the combination of the boxes in the block diagram and / or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0091] The units described in some embodiments of the present disclosure may be implemented in software or in hardware. The described units may also be provided in a processor, for example, may be described as: a processor including a splitting unit, a clustering unit, a preference unit, a generation unit, and a push unit. The names of these units do not, in some cases, constitute a limitation on the unit itself, for example, the splitting unit may also be described as "a unit that splits the feedback user feature vector set corresponding to the feedback target into each first-level feedback user feature vector subset according to preset time granularity information".

[0092] The functions described above herein may be performed, at least in part, by one or more hardware logic components. For example, and 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 chip (SOCs), complex programmable logic devices (CPLDs), and the like.

[0093] The above description is only an illustration of some preferred embodiments of the present disclosure and 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 the 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 (but not limited to) technical features with similar functions disclosed in the embodiments of the present disclosure.

Claims

1. An information push method, comprising: In response to receiving target tracking log data corresponding to a feedback target, the target tracking log data is parsed and processed to obtain a feedback user feature vector set, wherein the target tracking log data is log data received based on tracking of a target user behavior corresponding to an application page, the application page and the target user behavior correspond to the feedback target, the feedback target is to increase the click exposure rate of the application page, and the target user behavior refers to the click behavior of a user clicking to enter the application page; According to the preset time granularity information, the feedback user feature vector set corresponding to the feedback target is split into various first-level feedback user feature vector subsets, where each first-level feedback user feature vector subset corresponds to a time period, and the feedback target is the target of the user's behavior in the front-end application page; For each first-level feedback user feature vector subset in the first-level feedback user feature vector subsets, clustering the first-level feedback user feature vectors in the first-level feedback user feature vector subsets to obtain second-level feedback user feature vector subsets; Performing secondary feedback user feature vector subset optimization processing on each obtained secondary feedback user feature vector subset, and obtaining the secondary feedback user feature vector subset after optimization processing as the seed user feature vector set; In response to receiving a target user feature vector, generating similarity information based on the seed user feature vector set and the target user feature vector, wherein the target user feature vector is a feature vector of a user who visits the application page of the feedback target in real time; In response to the similarity information satisfying a preset similarity condition, the preset channel information corresponding to the feedback target is pushed to a user terminal corresponding to the target user feature vector.

2. The method according to claim 1, wherein The method further comprises: In response to the similarity information satisfying the preset similarity condition, the target user feature vector is added to the seed user feature vector set.

3. The method according to claim 1, wherein The generating of similarity information includes: performing averaging processing on each seed user feature vector in the seed user feature vector set to obtain a mean seed user feature vector; The similarity between the target user feature vector and the averaged seed user feature vector is determined as similarity information.

4. The method according to claim 1, wherein The parsing and processing of the target tracking log data includes: Performing field parsing on the target tracking log data to obtain a parsed user feature vector set; In response to the presence of identical analyzed user feature vectors in the analyzed user feature vector set, performing deduplication processing on the analyzed user feature vector set to obtain a deduplicated analyzed user feature vector set as a feedback user feature vector set; In response to the respective analyzed user feature vectors in the analyzed user feature vector set being different, the analyzed user feature vector set is determined as a feedback user feature vector set.

5. The method according to claim 1, wherein The method further comprises: In response to a feedback amount corresponding to a seed user feature vector in the seed user feature vector set detected within a preset time period meeting a preset feedback abnormality condition, deleting the seed user feature vector corresponding to the feedback amount from the seed user feature vector set, wherein the feedback amount corresponds to the feedback target.

6. An information push device, comprising: a parsing unit configured to, in response to receiving target embedding log data corresponding to a feedback target, parse the target embedding log data to obtain a feedback user feature vector set, wherein the target embedding log data is log data received based on embedding of a target user behavior corresponding to an application page, the application page and the target user behavior correspond to the feedback target, the feedback target is to increase the click exposure rate of the application page, and the target user behavior refers to the click behavior of a user clicking to enter the application page; a splitting unit configured to split the feedback user feature vector set corresponding to the feedback target into first-level feedback user feature vector subsets according to preset time granularity information, wherein each first-level feedback user feature vector subset corresponds to a time period, and the feedback target is a target corresponding to the user's behavior in the front-end application page; a clustering unit configured to perform clustering processing on each first-level feedback user feature vector subset in each first-level feedback user feature vector subset, to obtain each second-level feedback user feature vector subset; a selecting unit configured to perform a secondary feedback user feature vector subset selecting process on each obtained secondary feedback user feature vector subset, and obtain the secondary feedback user feature vector subset after the selecting process as a seed user feature vector set; a generating unit configured to generate similarity information based on the seed user feature vector set and the target user feature vector in response to receiving a target user feature vector, wherein the target user feature vector is a feature vector of a user who visits the application page of the feedback target in real time; The pushing unit is configured to push the preset channel information corresponding to the feedback target to the user terminal corresponding to the target user feature vector in response to the similarity information satisfying the preset similarity condition.

7. The device according to claim 6, wherein The device further comprises: An adding unit is configured to add the target user feature vector to the seed user feature vector set in response to the similarity information satisfying the preset similarity condition.

8. 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 according to any one of claims 1 to 5.

9. A computer-readable medium having a computer program stored thereon, wherein: When the program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images