Information Push Method, Device, Electronic Device and Storage Medium

Abstract

The present application relates to an information push method, device, electronic device and storage medium. Through the technical solution provided by the embodiment of the present application, when making information recommendations, the first interest feature of the target account for the target object in the target time period is determined. This first interest feature combines the browsing information of multiple accounts in the account set for the target object, and the browsing information of accounts other than the target account for the target object in the multiple accounts can be regarded as a baseline. The first interest feature determined based on this baseline can more accurately reflect the degree of interest of the target account in the target object. When determining the interest information, the time difference and the interest decay parameter are combined, and the obtained interest information can more accurately reflect the degree of interest of the target account in the target object at the current moment. Pushing information related to the target object to accounts whose interest information meets the target conditions also improves the accuracy of information push.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to an information push method, device, electronic device, and storage medium. Background Art

[0002] With the rapid development of network technology, more and more people are entertaining and relaxing through various information, including videos, articles, live broadcast links, product links, etc. Therefore, how to recommend information of interest to users is attracting more and more attention from information providers.

[0003] In related technologies, information provision platforms often recommend information of interest to users based on the users' followings. This means they recommend information related to the followings. For example, if a user follows celebrity A, the information provision platform can push information related to celebrity A to the user.

[0004] However, since a user may pay attention to multiple objects, there may be objects among these multiple objects that the user was interested in before but is not interested in now. When using the above method to recommend information, information that the user is not interested in may be recommended to the user. In other words, the accuracy of the information recommendation is not high and the information recommendation effect is not good. Summary of the Invention

[0005] This application provides an information push method, device, electronic device, and storage medium that can improve the accuracy of information recommendations and enhance the recommendation effect. The technical solution of this application is as follows:

[0006] On the one hand, a method for pushing information is provided, comprising:

[0007] determining, based on browsing information of multiple accounts in the account set for the target object within a target time period, a first interest characteristic of the target account in the target object within the target time period, the first interest characteristic being used to indicate a degree of interest of the target account in the target object, and the target account belonging to the account set;

[0008] Determining, based on the time difference between the target time period and the current moment, an interest decay parameter, and the first interest feature, interest information of the target account in the target object, wherein the interest information is used to indicate the degree of interest of the target account in the target object at the current moment, and the interest decay parameter is used to indicate the degree to which the degree of interest decays over time;

[0009] When the interest information meets the target condition, information related to the target object is pushed to the target account.

[0010] In a possible implementation, determining the first interest characteristic of the target account in the target object within the target time period based on browsing information of multiple accounts in the account set for the target object within the target time period includes:

[0011] Obtaining a first browsing count and a second browsing count corresponding to the target account within the target time period, where the first browsing count is the number of browsing times the target account has for the target object, and the second browsing count is the number of browsing times the target account has for objects in the object set to which the target object belongs;

[0012] Obtaining a first number of accounts and a second number of accounts corresponding to the target time period, wherein the first number of accounts is the number of accounts in the account set that have browsed any object in the object set during the target time period, and the second number of accounts is the number of accounts in the account set that have browsed the target object during the target time period;

[0013] The first interest feature is determined based on the first browsing count, the second browsing count, the number of the first accounts, and the number of the second accounts.

[0014] In a possible implementation, determining the first interest feature based on the first browsing count, the second browsing count, the number of the first accounts, and the number of the second accounts includes:

[0015] determining the quotient of the first browsing number and the second browsing number as a first browsing frequency of the target object by the target account;

[0016] determining a quotient of the number of the first accounts and the number of the second accounts as a second browsing frequency of the target object by the multiple accounts;

[0017] The product of the first browsing frequency and the second browsing frequency is determined as the first interest feature.

[0018] In a possible implementation, determining the target account's interest information for the target object based on the time difference between the target time period and the current moment, the interest decay parameter, and the first interest feature includes:

[0019] Determining a time decay value of the target account to the target object based on the time difference and the interest decay parameter, wherein the time decay value is negatively correlated with the time difference;

[0020] determining a product of the first interest feature and the time decay value as a second interest feature, where the second interest feature is used to indicate the degree of interest of the target account in the target object at the current moment;

[0021] The second interest feature is normalized to obtain the interest information.

[0022] In a possible implementation, there are multiple target time periods, and determining the first interest characteristic of the target account in the target object within the target time period based on browsing information of multiple accounts in the account set for the target object within the target time period includes:

[0023] determining, based on browsing information of a plurality of accounts in the account set for a target object within a plurality of target time periods, a plurality of first interest features of the target account for the target object within the plurality of time periods;

[0024] The determining, based on the time difference between the target time period and the current time, the interest decay parameter, and the first interest feature, the target account's interest information for the target object includes:

[0025] Based on multiple time differences between the multiple target time periods and the current moment, the interest decay parameter, and the multiple first interest features, the interest information of the target account for the target object is determined.

[0026] In a possible implementation, determining the target account's interest information for the target object based on multiple time differences between the multiple target time periods and the current moment, the interest decay parameter, and the multiple first interest features includes:

[0027] Determining an initial second interest feature corresponding to each target time period based on a time difference between each target time period and the current moment, the interest decay parameter, and the first interest feature corresponding to each target time period;

[0028] Based on the initial second interest features corresponding to each of the target time periods, the interest information of the target account in the target object is obtained.

[0029] In a possible implementation, obtaining the target account's interest information for the target object based on the initial second interest feature corresponding to each target time period includes:

[0030] Fusion of the initial second interest features corresponding to the target time periods to obtain the second interest features of the target account for the target object;

[0031] The second interest feature is normalized to obtain the interest information of the target account in the target object.

[0032] In a possible implementation, when the interest information meets the target condition, pushing information related to the target object to the target account includes any of the following:

[0033] When the interest information is greater than or equal to the interest information threshold, pushing information related to the target object to the target account;

[0034] In the case where the interest information is the first N interest information among multiple interest information, information related to the target object is pushed to the target account, N is a positive integer, and the multiple interest information is the interest information of multiple accounts in the account set on the target object at the current moment.

[0035] In a possible implementation manner, the time difference between the target time period and the current time is any one of the following:

[0036] The difference between the earliest time in the target time period and the current time;

[0037] The difference between the midpoint of the target time period and the current time.

[0038] On the one hand, an information push device is provided, comprising:

[0039] a first interest characteristic determining unit configured to determine, based on browsing information of a plurality of accounts in the account set for the target object within a target time period, a first interest characteristic of a target account in the target object within the target time period, wherein the first interest characteristic is used to indicate a degree of interest of the target account in the target object, and the target account belongs to the account set;

[0040] an interest information determining unit, configured to determine interest information of the target account in the target object based on a time difference between the target time period and the current moment, an interest decay parameter, and the first interest feature, wherein the interest information is used to indicate a degree of interest of the target account in the target object at the current moment, and the interest decay parameter is used to indicate a degree of decay of the interest over time;

[0041] The push unit is configured to push information related to the target object to the target account when the interest information meets the target condition.

[0042] In a possible embodiment, the first interest feature determination unit is configured to execute acquisition of a first browsing count and a second browsing count corresponding to the target account within the target time period, where the first browsing count is the browsing count of the target account for the target object, and the second browsing count is the browsing count of the target account for objects in the object set to which the target object belongs, the object set including multiple objects, and the target object belongs to the object set; acquisition of a first account quantity and a second account quantity corresponding to the target time period, where the first account quantity is the number of accounts in the account set that have browsed any object in the object set within the target time period, and the second account quantity is the number of accounts in the account set that have browsed the target object within the target time period; and determination of the first interest feature based on the first browsing count, the second browsing count, the first account quantity, and the second account quantity.

[0043] In one possible implementation, the first interest feature determination unit is configured to determine the quotient of the first browsing number and the second browsing number as the first browsing frequency of the target account for the target object; determine the quotient of the number of the first accounts and the number of the second accounts as the second browsing frequency of the target object by the multiple accounts; and determine the product of the first browsing frequency and the second browsing frequency as the first interest feature.

[0044] In one possible implementation, the interest information determination unit is configured to determine a time decay value of the target account's interest in the target object based on the time difference and the interest decay parameter, wherein the time decay value is negatively correlated with the time difference; determine a second interest feature as a product of the first interest feature and the time decay value, wherein the second interest feature is used to represent the target account's level of interest in the target object at the current moment; normalize the second interest feature to obtain the interest information; and normalize the second interest feature to obtain the interest value.

[0045] In one possible embodiment, there are multiple target time periods, and the first interest feature determination unit is configured to determine the multiple first interest features of the target account for the target object in the multiple time periods based on the browsing information based on the target object of multiple accounts in the account set in the multiple target time periods; the interest information determination unit is configured to determine the interest information of the target account for the target object based on multiple time differences between the multiple target time periods and the current moment, the interest decay parameters and the multiple first interest features.

[0046] In a possible embodiment, the first interest feature determination unit is configured to determine the initial second interest feature corresponding to each target time period based on the time difference between each target time period and the current moment, the interest decay parameter, and the first interest feature corresponding to each target time period; and obtain the interest information of the target account for the target object based on the initial second interest feature corresponding to each target time period.

[0047] In a possible implementation, the first interest feature determination unit is configured to perform a fusion of the initial second interest features corresponding to each of the target time periods to obtain the second interest features of the target account for the target object; and normalize the second interest features to obtain the interest information of the target account for the target object.

[0048] In a possible implementation, the pushing unit is configured to perform any of the following:

[0049] When the interest information is greater than or equal to the interest information threshold, pushing information related to the target object to the target account;

[0050] In the case where the interest information is the first N interest information among multiple interest information, information related to the target object is pushed to the target account, N is a positive integer, and the multiple interest information is the interest information of multiple accounts in the account set on the target object at the current moment.

[0051] In a possible implementation manner, the time difference between the target time period and the current time is any one of the following:

[0052] The difference between the earliest time in the target time period and the current time;

[0053] The difference between the midpoint of the target time period and the current time.

[0054] In one aspect, an electronic device is provided, comprising:

[0055] processor;

[0056] a memory for storing instructions executable by the processor;

[0057] The processor is configured to execute the instructions to implement the above-mentioned information pushing method.

[0058] On the one hand, a computer-readable storage medium is provided. When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is capable of executing the above-mentioned information push method.

[0059] In one aspect, a computer program product is provided, comprising a computer program, which implements the above-mentioned information push method when executed by a processor.

[0060] The technical solutions provided by the embodiments of this application bring at least the following beneficial effects:

[0061] Through the technical solution provided by the embodiment of the present application, when making information recommendations, the first interest feature of the target account for the target object in the target time period is determined. This first interest feature combines the browsing information of the target object by multiple accounts in the account set, and the browsing information of the target object by accounts other than the target account in the multiple accounts can be regarded as a baseline. The first interest feature determined based on this baseline can more accurately reflect the target account's interest in the target object. When determining the interest information, the time difference and the interest decay parameter are combined, that is, the influence of time on the interest level is taken into account, and the obtained interest information can more accurately reflect the target account's interest level in the target object at the current moment. Pushing information related to the target object to accounts whose interest information meets the target conditions also improves the accuracy of information push.

[0062] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the specification are used to explain the principles of the present application, and do not constitute an improper limitation on the present application.

[0064] Figure 1 is a schematic diagram of a forgetting curve according to an exemplary embodiment.

[0065] Figure 2 The figure is a schematic diagram of an implementation environment of an information push method according to an exemplary embodiment.

[0066] Figure 3 The figure is a flowchart of an information push method according to an exemplary embodiment.

[0067] Figure 4 The figure is a flowchart of an information push method according to an exemplary embodiment.

[0068] Figure 5 is a schematic diagram showing a decay curve of the interest level according to an exemplary embodiment.

[0069] Figure 6is a schematic diagram showing a decay curve of the interest level according to an exemplary embodiment.

[0070] Figure 7 The figure is a flowchart of an information push method according to an exemplary embodiment.

[0071] Figure 8 The figure is a block diagram of an information pushing device according to an exemplary embodiment.

[0072] Figure 9 The figure is a block diagram of a server according to an exemplary embodiment. DETAILED DESCRIPTION

[0073] In order to enable ordinary people in the art to better understand the technical solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0074] It should be noted that the terms "first," "second," and the like in the specification and claims of this application and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate so that the embodiments of the application described herein can be implemented in an order other than those illustrated or described herein. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0075] The user information involved in this application may be information authorized by the user or fully authorized by all parties.

[0076] The following is an introduction to some terms in the embodiments of the present application:

[0077] TF-IDF (Term Frequency-Inverse Document Frequency) algorithm: is a commonly used weighting technology for information retrieval and data mining. TF-IDF is a statistical method used to evaluate the importance of a word to a document set or one of the documents in a corpus. The importance of a word increases in direct proportion to the number of times it appears in a document, but at the same time decreases in inverse proportion to the frequency of its appearance in the corpus. The main idea of ​​TFFIDF is: if a word or phrase appears in an article with a high frequency TF and rarely appears in other articles, it is considered that this word or phrase has good category discrimination ability and is suitable for classification. TF represents the frequency of the term in document d. IDF means: if the fewer documents containing the term t, that is, the smaller n is, the larger the IDF is, it means that the term t has good category discrimination ability. If the number of documents in a certain category C that contain the term t is m, and the total number of documents in other categories that contain t is k, then it is obvious that the total number of documents containing t is n=m+k. When m is large, n is also large, and the IDF value obtained according to the IDF formula will be small, which means that the term t has weak ability to distinguish categories.

[0078] The forgetting curve was discovered by German psychologist Hermann Ebbinghaus. It describes the pattern of how the human brain forgets new things. People can learn and utilize the forgetting pattern from the forgetting curve to improve their memory. The forgetting curve tells us that forgetting in learning is regular, the process of forgetting is fast, and it starts fast and then slows down. For an example of the forgetting curve, see Figure 1 The time decay algorithm is based on the idea of the forgetting curve.

[0079] Figure 2 This is a schematic diagram of an implementation environment of an information push method provided in an embodiment of the present application, see Figure 2 , the implementation environment includes a terminal 201 and a server 202.

[0080] The terminal 201 may be at least one of a smartphone, a smartwatch, a desktop computer, a laptop computer, and the like. An application supporting information display may be installed and run on the terminal 201. A user may log in to the application through the terminal 201 and view information through the application. The information may include videos, articles, links to live broadcasts, and shopping links. In some embodiments, the application may have a user account logged in.

[0081] Terminal 201 may generally refer to one of multiple terminals. This embodiment uses terminal 201 as an example. Those skilled in the art will appreciate that the number of terminals may be greater or lesser. For example, there may be only a few terminals 201, or there may be dozens, hundreds, or even more terminals 201. This embodiment of the application does not limit the number or device type of terminals 201. Terminal 201 may be connected to server 202 via a wireless network or a wired network.

[0082] The server 202 may be at least one of a single server, multiple servers, a cloud computing platform, and a virtualization center. The server 202 provides background services for applications running on the terminal 201, such as recommending information to the application.

[0083] In some embodiments, the number of the servers 202 may be more or less, which is not limited in the present embodiment. Of course, the servers 202 may also include other functional servers to provide more comprehensive and diversified services.

[0084] After introducing the implementation environment of the embodiment of the present application, the application scenario of the information push method provided by the embodiment of the present application will be described in combination with the above-mentioned implementation environment. In the following description process, the server is the server 202 in the above-mentioned implementation environment, and the terminal is the terminal 201 in the above-mentioned implementation environment.

[0085] The information push method provided in the embodiment of the present application can be applied in scenarios of live broadcast recommendation, product recommendation, video recommendation and article recommendation.

[0086] In the scenario of live broadcast recommendation, the information recommendation method provided by the embodiment of the present application can determine whether to recommend the live broadcast room to the target account in the account set based on the browsing information of multiple accounts in the account set for the live broadcast room host. The target account is also the user account to be recommended for the live broadcast room, wherein the account set includes all user accounts on the information providing platform, or the account set is an account set divided by technical personnel. For example, after the server obtains the browsing information of the live broadcast room host based on multiple accounts in the account set, it can determine the first interest feature of the target account for the host within the target time period based on the browsing information. The first interest feature can indicate the target account's level of interest in the host, wherein the target time period is any time period before the current moment, such as from 90 days to 60 days from the current moment. The server determines the target account's interest information for the host based on the time difference between the target time period and the current moment, the interest decay parameter, and the first interest feature. The interest information can indicate the target account's level of interest in the host at the current moment. The server can push the anchor's live broadcast room to the target account based on the target account's interest information. For example, the server can push the link of the live broadcast room to the target account, or push reminder information for the anchor to the target account when the anchor starts broadcasting.

[0087] In the context of product recommendation, the information recommendation method provided in the embodiment of the present application can determine whether to recommend a product to a target account in the account set based on the browsing information of multiple accounts in the account set for the product. The target account is also the user account to be recommended for the product, wherein the account set includes all user accounts on the information provision platform, or the account set is an account set divided by technical personnel. For example, after the server obtains the browsing information of the product based on multiple accounts in the account set, it can determine the first interest feature of the target account for the product within the target time period based on the browsing information. The first interest feature can indicate the target account's level of interest in the product, wherein the target time period is any time period before the current moment, such as from 90 days to 60 days from the current moment. The server determines the target account's interest information for the product based on the time difference between the target time period and the current moment, the interest decay parameter, and the first interest feature. The interest information can indicate the target account's level of interest in the product at the current moment. The server can push the product of the product to the target account based on the interest information of the target account. For example, the server can push the link of the product to the target account, or push reminder information about the product to the target account when there is a promotion for the product.

[0088] In the scenario of video recommendation, the information recommendation method provided by the embodiment of the present application can determine whether to recommend the video to the target account in the account set based on the browsing information of multiple accounts in the account set for the video author. The target account is also the user account to be recommended for the video, wherein the account set includes all user accounts on the information providing platform, or the account set is an account set divided by technical personnel. For example, after the server obtains the browsing information of the video author based on multiple accounts in the account set, it can determine the first interest feature of the target account in the video author within the target time period based on the browsing information. The first interest feature can represent the target account's interest level in the video author, wherein the target time period is any time period before the current moment, such as from 90 days to 60 days from the current moment. The server determines the target account's interest information in the video author based on the time difference between the target time period and the current moment, the interest decay parameter, and the first interest feature. The interest information can represent the target account's interest level in the video author at the current moment. The server can push the video of the video author to the target account based on the interest information of the target account. For example, the server can push the link of the video to the target account, or push reminder information for the video author to the target account when the video author updates the video.

[0089] In the context of article recommendation, the information recommendation method provided by the embodiment of the present application can determine whether to recommend the article to a target account in the account set based on the browsing information of multiple accounts in the account set for the article author. The target account is also the user account to be recommended for the article, wherein the account set includes all user accounts on the information provision platform, or the account set is an account set divided by technical personnel. For example, after the server obtains the browsing information of the article author based on multiple accounts in the account set, it can determine the first interest feature of the target account in the article author within the target time period based on the browsing information. The first interest feature can represent the target account's level of interest in the article author, wherein the target time period is any time period before the current moment, such as from 90 days to 60 days from the current moment. The server determines the target account's interest information in the article author based on the time difference between the target time period and the current moment, the interest decay parameter, and the first interest feature. The interest information can represent the target account's level of interest in the article author at the current moment. The server can push the article of the author of the article to the target account based on the interest information of the target account. For example, the server can push the link of the article to the target account, or push reminder information for the author of the article to the target account when the author of the article updates the article.

[0090] It should be noted that in the above description, the information push method provided in the embodiment is used as an example to illustrate that it can be applied to the scenarios of live broadcast recommendation, product recommendation, video recommendation and article recommendation. In other possible implementation methods, the information push method provided in the embodiment of the present application can also be applied to recommendation scenarios of other types of information, and the embodiment of the present application does not limit this.

[0091] After introducing the implementation environment and application scenarios of the embodiment of the present application, the information recommendation method provided by the embodiment of the present application is described below. Figure 3 , taking the execution subject as a server as an example, the method includes:

[0092] 301. The server determines a first interest feature of the target account for the target object within the target time period based on browsing information of multiple accounts in the account set for the target object within the target time period. The first interest feature is used to indicate the degree of interest of the target account in the target object. The target account belongs to the account set.

[0093] In some embodiments, the multiple accounts in the account set are all accounts on the information providing platform, or are accounts in the account set divided by technicians based on actual conditions, which is not limited in this embodiment of the present application. The target time period is the time period before the current moment. The browsing information for the target object includes browsing the target object's news, videos, live broadcast room, and other information related to the target object.

[0094] 302. The server determines the target account's interest information for the target object based on the time difference between the target time period and the current moment, the interest decay parameter, and the first interest feature. The interest information is used to indicate the target account's interest level in the target object at the current moment, and the interest decay parameter is used to indicate the degree of decay of the interest level over time.

[0095] The interest decay parameter can be used to indicate the degree to which the interest level decreases over time. In some embodiments, the larger the interest decay parameter, the greater the decrease in interest level per unit time. In some embodiments, the interest decay parameter is set by a technician based on actual conditions, such as 1, 2, or other numbers.

[0096] 303. When the interest information meets the target condition, the server pushes information related to the target object to the target account.

[0097] Among them, the information related to the target object includes the live broadcast room link, video, article and other information related to the target object, which is not limited in the embodiment of the present application.

[0098] Through the technical solution provided by the embodiment of the present application, when making information recommendations, the first interest feature of the target account for the target object in the target time period is determined. This first interest feature combines the browsing information of the target object by multiple accounts in the account set, and the browsing information of the target object by accounts other than the target account in the multiple accounts can be regarded as a baseline. The first interest feature determined based on this baseline can more accurately reflect the target account's interest in the target object. When determining the interest information, the time difference and the interest decay parameter are combined, that is, the influence of time on the interest level is taken into account, and the obtained interest information can more accurately reflect the target account's interest level in the target object at the current moment. Pushing information related to the target object to accounts whose interest information meets the target conditions also improves the accuracy of information push.

[0099] The above steps 301-303 are a brief description of the information push method provided by the embodiment of the present application. The following will combine some examples to provide a more detailed description of the information push method provided by the embodiment of the present application. Figure 4 , taking the execution subject as a server as an example, the method includes:

[0100] 401. The server determines the target time period.

[0101] The target time period is a historical time period during which the server determines the target account's interest in the target object. The target time period may be one or more, and this embodiment of the application does not limit this.

[0102] In a possible implementation, the server determines a target object and obtains the target time period corresponding to the target object, wherein information related to the target object is information to be recommended.

[0103] In this embodiment, since the information pushed by the server is information related to the target object, the server can determine the target time period based on the target object, and the determination of the target time period is more in line with the needs of the target object.

[0104] For example, the server obtains an information recommendation request, which carries the target time period, the identifier of the target object, and information related to the target object, wherein the information related to the target object is also the information to be pushed. The information recommendation request is sent to the server by the terminal corresponding to the target object. The terminal corresponding to the target object may refer to the terminal corresponding to the user account of the target object, or to the terminal corresponding to other user accounts associated with the target object. This embodiment of the present application does not limit this. The server obtains the target time period, the identifier of the target object, and information related to the target object from the information recommendation request.

[0105] For example, if the target object is a user, then when the user needs to have information related to them pushed to the server, they can send an information recommendation request to the server via the terminal. In some embodiments, the terminal displays an information recommendation page that is used to obtain the user's identifier, the target time period, and information related to the user. The user can fill in the corresponding content on the information recommendation page. In response to an operation on the information recommendation page, the terminal sends an information recommendation request to the server. The information recommendation request carries the content of the information recommendation page, that is, the information recommendation request carries the user's identifier, the target time period, and information related to the user. After receiving the information recommendation request, the server obtains the user's identifier, the target time period, and information related to the user from the information recommendation request. If there are multiple target time periods, the terminal can obtain multiple target time periods through the information recommendation page. That is, the user enters the multiple target time periods on the information recommendation page, and the terminal sends the multiple target time periods to the server, which then obtains the multiple target time periods. Alternatively, after the terminal obtains the time period entered by the user through the information recommendation page, it sends the time period to the server, and the server divides the time period into multiple target time periods.

[0106] Taking the scenario where the information push method provided in the embodiments of the present application is used for live broadcast recommendation as an example, the target object is the host of the live broadcast room. In some embodiments, the host can enter their identity document (ID), the target time period, and the link of the live broadcast room they want to push on the information recommendation page displayed on the terminal. The terminal sends the ID, target time period, and link of the live broadcast room entered by the host on the information recommendation page to the server, and the server obtains the ID, target time period, and link of the live broadcast room.

[0107] In some embodiments, the target time period is configured by technical personnel according to actual conditions, for example, the target time period is configured to be 30 days, 60 days or 90 days between the current time, etc., and this embodiment of the present application does not limit this.

[0108] 402. The server determines the first interest feature of the target account for the target object within the target time period based on browsing information of multiple accounts in the account set for the target object within the target time period. The first interest feature is used to indicate the degree of interest of the target account in the target object. The target account belongs to the account set.

[0109] In one possible implementation, the server obtains a first and second browsing count corresponding to the target account within the target time period, where the first browsing count is the target account's browsing count for the target object, and the second browsing count is the target account's browsing count for objects in the object set to which the target object belongs. The server obtains a first number of accounts and a second number of accounts corresponding to the target time period, where the first number of accounts is the number of accounts in the account set that have browsed any object in the object set within the target time period, and the second number of accounts is the number of accounts in the account set that have browsed the target object within the target time period. The server determines the first interest feature based on the first and second browsing counts, the first and second number of accounts.

[0110] The number of views for the target object refers to the number of views of the relevant information of the target object. The relevant information of the target object here includes the target object's videos, articles, and live broadcast rooms. That is, the number of views for the target object includes the number of views of the videos, articles, and live broadcast rooms related to the target object. In some embodiments, the number of views for the target object also includes the number of searches for the target object. The object set to which the target object belongs includes multiple objects, and the multiple objects are objects of the same type. If the target object is a celebrity, then all the objects in the object set are celebrities; if the target object is a writer, then all the objects in the object set are writers; if the target object is a television, then all the objects in the object set are televisions. In some embodiments, the division of the object set is to classify the objects. The division of the object set is performed by technical personnel based on actual conditions and is not limited in the embodiments of the present application. The account set is the account set to which the target account belongs. In some embodiments, the account set is all accounts on the information provision platform, or the account set is the account set to which the relevant information of the target object is to be pushed. The embodiments of the present application do not limit this.

[0111] Under this embodiment, the server can determine the first interest feature based on the first number of views, the second number of views, the number of first accounts and the number of second accounts. That is, when determining the first interest feature, it not only refers to the number of views of the target account on the target within the target time period, but also combines the number of views of the target account on the object set to which the target object belongs during the target time period and the number of first accounts and second accounts in the account set that have browsed the target object and the object set during the target time period. The determined first account feature can more accurately reflect the target account's interest in the target object.

[0112] In order to explain the above embodiment more clearly, the following will be divided into several parts to explain the above embodiment.

[0113] In the first part, the server obtains the first browsing count and the second browsing count corresponding to the target account within the target time period.

[0114] In one possible implementation, the server queries a corresponding maintained database based on the target account and the target time period to obtain the first number of views and the second number of views corresponding to the target account within the target time period, wherein the database stores multiple accounts and the first number of views and the second number of views corresponding to the multiple accounts, respectively.

[0115] For example, the server generates a first query request based on the target account, the target time period, and the identifier of the target object. The server queries the database based on the first query request to obtain a first number of views corresponding to the target account, which includes the number of views of the target account on videos, articles, and live broadcast rooms related to the target object. The server determines the object set to which the target object belongs based on the identifier of the target object. The server generates a second query request based on the target account, the target time period, and the identifier of the object set, and queries the database based on the second query request to obtain a second number of views corresponding to the target account, which includes the number of views of the target account on videos, articles, and live broadcast rooms related to the objects in the account set.

[0116] Taking the target object as a celebrity, for example, the server generates a first query request based on the target account, the target time period, and the celebrity's identifier. This first query request is written in Structured Query Language (SQL). Based on the first query request, the server queries the database and obtains a first view count corresponding to the target account. This first view count is the number of views the target account has made on the celebrity, including views of videos and articles related to the celebrity, as well as views of the celebrity's livestream. Based on the celebrity's identifier, the server determines the object set to which the celebrity belongs, where all objects in the object set are celebrities. Based on the target account, the target time period, and the identifier of the object set, the server generates a second query request written in Structured Query Language (SQL). Based on the second query request, the server queries the database and obtains a second view count corresponding to the target account. This second view count is the number of views the target account has made on the object set, or more specifically, on the celebrities in the object set.

[0117] In addition, when the target object is a writer or a commodity, the corresponding object set is a writer set or a commodity set respectively. The implementation process belongs to the same inventive concept as the above description and will not be repeated here.

[0118] In one possible implementation, the server obtains the target account's historical browsing history and, based on the target object's identifier and the target time period, obtains a first number of views of the target object from the historical browsing history. Based on the target object's identifier, the server determines the object set to which the target object belongs. Based on the object set, the server obtains a second number of views of the target account from the historical browsing history.

[0119] For example, the server queries the corresponding browsing record database based on the target account to obtain the historical browsing record of the target account. The browsing record database stores multiple accounts and the historical browsing records corresponding to the multiple accounts. The server filters the historical browsing record based on the target time period to obtain the target browsing record of the target account within the target time period, or, when obtaining the historical browsing record of the target account from the browsing record database, it queries based on the target time period to obtain the target browsing record. The server determines the first number of browsings of the target account for the target object based on the identifier of the target object and the target browsing record. The server determines the object set to which the target object belongs based on the identifier of the target object. The server determines the second number of browsings of the target account for the object set based on the identifiers of multiple objects in the object set and the target browsing record.

[0120] Continuing with the example of a celebrity as the target object, the server retrieves the historical browsing records corresponding to the target account from the browsing history database. The server filters the historical browsing records based on the target time period to obtain the target browsing records. The server counts the first number of views of the celebrity by the target account during the target time period based on the celebrity's identifier, such as the celebrity's name or stage name. The server determines the object set to which the celebrity belongs, and counts the target browsing records based on multiple objects in the object set to obtain the second number of views of the target account for the multiple objects in the object set.

[0121] In the second part, the server obtains the number of first accounts and the number of second accounts corresponding to the target time period.

[0122] In a possible implementation, the server queries a corresponding maintained database based on the target account, the target time period, and the identifier of the target object to obtain the first account quantity and the second account quantity corresponding to the target time period.

[0123] For example, based on the target account, the server determines the account set to which the target account belongs. Based on the account set, the target time period, and the identifier of the target object, the server determines a second number of accounts in the account set that browsed the target object during the target time period, where browsing behavior on the target object includes clicking on videos or articles related to the target object, and watching live streams corresponding to the target object. Based on the identifier of the target object, the server determines the object set to which the target object belongs. The server determines a first number of accounts in the account set that browsed any object in the object set during the target time period. When determining the first number of accounts, any account in the account set that browsed any object in the object set is counted in the first number of accounts. If an account in the object set browsed any object in the object set multiple times during the target time period, or if the account browsed multiple objects in the object set during the target time period, the account is counted only once. In other words, the first number of accounts is not related to the number of views by an individual account. When determining the second account count, any account in the account set that has browsed the target object will be counted in the second account count. If an account in the object set has browsed the target object multiple times during the target time period, the account will only be counted once. In other words, the second account count is not related to the number of views by a single account.

[0124] Taking the target object as a celebrity as an example, the server determines the account set to which the target account belongs based on the target account. Based on the account set, the target time period, and the celebrity's identifier, the server determines the second number of accounts in the account set that have browsed the celebrity during the target time period. If the account set is all accounts on the information provision platform, the second number of accounts is the number of accounts on the information provision platform that have browsed the celebrity during the target time period. The server determines the object set to which the celebrity belongs, where multiple objects in the object set are celebrities. The server determines the first number of accounts in the account set that have browsed any object in the object set during the target time period. If the account set is all accounts on the information provision platform, the first number of accounts is the number of accounts on the information provision platform that have browsed the celebrity set during the target time period.

[0125] Part 3: The server determines the first interest feature based on the first browsing number, the second browsing number, the number of the first accounts, and the number of the second accounts.

[0126] In one possible implementation, the server determines the quotient of the first number of views and the second number of views as the first viewing frequency of the target account for the target object. The server determines the quotient of the number of first accounts and the number of second accounts as the second viewing frequency of the target object by the multiple accounts. The server determines the product of the first viewing frequency and the second viewing frequency as the first interest feature.

[0127] Under this implementation, since the first browsing frequency is the quotient of the first browsing number and the second browsing number, and the first browsing number is the browsing number of the target account for the target object, and the second browsing number is the browsing number of the target account for the object set to which the target object belongs, then the first browsing frequency can reflect the target account's attention to the target object within the target time period. The higher the first browsing frequency, the higher the target account's attention to the target object compared to other objects in the object set; the lower the first browsing frequency, the lower the target account's attention to the target object compared to other objects in the object set. Since the second browsing frequency is the quotient of the first and second account numbers, and the first account number is the number of accounts in the account set that have browsed any object in the object set during the target time period, and the second account number is the number of accounts in the account set that have browsed the target object during the target time period, the second browsing frequency can reflect the browsing popularity of the target object from the account set dimension. The higher the second browsing frequency, the lower the browsing popularity of the target object during the target time period; the lower the second frequency, the higher the browsing popularity of the target object during the target time period. Through this fusion of the account dimension and the account set dimension, the first interest feature obtained can more accurately reflect the target account's interest in the target object.

[0128] For example, the server can obtain the first browsing frequency through the following formula (1), obtain the second browsing frequency through the following formula (2), and obtain the first interest feature based on the first browsing frequency and the second browsing frequency through the following formula (3).

[0129]

[0130] Among them, tf i，j is the first browsing frequency, n i，j is the first browsing number, ∑ k n k，j This is the second view count.

[0131]

[0132] Among them, idf iis the second browsing frequency, |D| is the number of first accounts, j is the number of second accounts, t i is the target object, d j A collection of objects.

[0133] tfidf i，j =tf i，j ×idf i (3)

[0134] Among them, tfidf i，j The first feature of interest.

[0135] If the target object is a celebrity (hereinafter referred to as A), when the server determines the target account's first interest feature in A, tf i，j = target account page views of A / page views of all celebrities in the account set, idf i = log(number of accounts in the account set that browse celebrities / (number of accounts that browse celebrity A)), tf i，j It can determine the popularity of A in the entire network during the target time period. The larger the value, the higher the popularity of A. i，j , reflecting the target account's level of interest in the target subject, based on overall online popularity. Since celebrities represent a unique demographic, their viewing is likely driven by public opinion, rather than positive influence. Some users watch celebrity-related content simply for the sake of excitement. By comprehensively considering the impact of celebrity popularity through the first interest feature, we can penalize user behavior during periods of celebrity popularity, more accurately reflecting user interest in the celebrity.

[0136] In a possible implementation, there are multiple target time periods, and the server determines multiple first interest features of the target account for the target object in the multiple target time periods based on the browsing information of multiple accounts in the account set for the target object in the multiple target time periods.

[0137] Among them, when the server determines multiple first interest characteristics of the target account for the target object within multiple target time periods, it belongs to the same inventive concept as the previously described method of determining the first interest characteristics of the target account for the target object within a target time period. The implementation process can be found in the previous description and will not be repeated here.

[0138] 403. The server determines the target account's interest information for the target object based on the time difference between the target time period and the current moment, the interest decay parameter, and the first interest feature. The interest information is used to indicate the target account's interest level in the target object at the current moment, and the interest decay parameter is used to indicate the degree of decay of the interest level over time.

[0139] In some embodiments, the time difference between the target time period and the current moment refers to the difference between the earliest moment in the target time period and the current moment, or the difference between the midpoint moment in the target time period and the current moment.

[0140] In one possible implementation, the server determines a time decay value of the target account for the target object based on the time difference and the interest decay parameter, where the time decay value is negatively correlated with the time difference. The server determines the product of the first interest feature and the time decay value as a second interest feature, where the second interest feature is used to represent the target account's level of interest in the target object at the current moment. The server normalizes the second interest feature to obtain the interest information.

[0141] In this embodiment, the server uses the interest decay parameter to represent the degree of interest decay over time, and processes the first interest feature in combination with the time difference to obtain the second interest feature. Since the first interest feature represents the target account's interest in the target object within the target time period, the second interest feature can reflect the target account's interest in the target object at the current moment after combining the interest decay parameter and the time difference. The interest information obtained by the server based on the second interest feature can also accurately reflect the target account's interest in the target object at the current moment.

[0142] For example, the server obtains the second interest feature through the following formula (4) and obtains the interest information through the following formula (5).

[0143] x=tfidf i ×e -γ·Δt (4)

[0144] Where x is the second feature of interest, γ is the decay parameter of interest, and Δt is the time difference.

[0145]

[0146] Among them, S is interest information.

[0147] In one possible implementation, there are multiple target time periods, and the server determines the target account's interest information for the target object based on multiple time differences between the multiple target time periods and the current moment, the interest decay parameter, and the multiple first interest features.

[0148] For example, the server determines the initial second interest feature corresponding to each target time period based on the time difference between each target time period and the current moment, the interest decay parameter, and the first interest feature corresponding to each target time period. Based on the initial second interest feature corresponding to each target time period, the server obtains the target account's interest information for the target object.

[0149] Among them, the method in which the server determines the initial second interest feature corresponding to each target time period based on the time difference between each target time period and the current moment, the interest decay parameter, and the first interest feature corresponding to each target time period belongs to the same inventive concept as the method for determining the second interest feature in the previous implementation. The implementation process is described above and will not be repeated here. The following describes the method in which the server obtains the target account's interest information for the target object based on the initial second interest feature corresponding to each target time period in the above example.

[0150] In a possible implementation, the server fuses the initial second interest features corresponding to each target time period to obtain the second interest features of the target account for the target object. The server normalizes the second interest features to obtain the interest information of the target account for the target object. Figure 5 , which shows the decay curve of the interest level of a target object in a single browsing behavior. In some embodiments, if a user purchased a certain product six months ago but has not taken any action on it since then (browsing, collecting, adding to shopping cart, and purchasing), then the user's interest decay curve for the product is as follows: Figure 5 As shown in the figure, γ is the interest decay parameter, t0, t1 and t2 are different moments, P0 is the interest level at t0, and P1 is the interest level at t1. As can be seen from the figure, the interest level decreases with the passage of time. Figure 6 , shows the interest decay curve when the target account performs repeated browsing behavior on the target object in different time periods. The beginning of each time period is when the target account has browsed the target object. γ is the interest decay parameter, t0, t1 and t2 are different moments, P0 is the interest level at t0, P1 is the interest level at t1, and h1 and h2 are the differences in interest levels.

[0151] For example, the server adds the initial second interest features corresponding to each target time period to obtain the second interest feature of the target account for the target object, and normalizes the second interest feature to obtain the interest information of the target account for the target object.

[0152] For example, the server adds the initial second interest features corresponding to each target time period using the following formula (6) to obtain the second interest feature of the target account for the target object. The server normalizes the second interest feature using the above formula (5) to obtain the interest information of the target account for the target object.

[0153]

[0154] Where i is the sequence number of the target time period, and n is the number of target time periods.

[0155] The above steps 402 and 403 are described below with an example.

[0156] In the following example, there are four target time periods, namely, 120 to 90 days from the current time, 90 to 60 days from the current time, 60 to 30 days from the current time, and 30 days before the current time.

[0157] The target account's browsing history for celebrity A is known:

[0158] The target account has visited A twice in the past 120 to 90 days (the first number of views), and visited all celebrities five times (the second number of views). The number of celebrity accounts that have browsed the entire site (account collection) is 100 (the first number of accounts), and the number of accounts that have browsed celebrity A is 80 (the second number of accounts).

[0159] The target account has visited A four times in the past 90 to 60 days and visited all stars six times. The number of accounts that visited stars on the entire site is 100, and the number of accounts that visited star A is 100.

[0160] The target account has not browsed A in the past 60 to 30 days.

[0161] The target account has visited A once in the past 30 days and visited all celebrities six times. The number of accounts that visited celebrities on the entire site is 100, and the number of accounts that visited celebrity A is 10.

[0162] Taking the data corresponding to the target time period from 120 days to 90 days as an example, the target account’s first interest feature tfidf for star A is i，j =(2 / 5)×log(100 / 80)=0.0338. Taking the interest decay parameter γ as 1 and the time difference as the difference between the earliest moment in the target time period and the current moment as an example, the target account's initial second interest feature for star A is x=0.0338×e -120For other target time periods, the above method can be used to calculate the corresponding initial second interest features. The server merges the initial second interest features of each target time period into a second interest feature, normalizes the second interest feature, and obtains the target account's interest information for star A.

[0163] The above steps 401-403 can also be completed by Figure 7 To indicate, see Figure 7 The server determines a time range, that is, a target time period. The server determines the target account's first interest characteristic for the target object. The server determines the target account's interest information for the target object. The server outputs the result.

[0164] 404. The server determines whether the interest information meets the target condition.

[0165] The target condition refers to whether the interest information is greater than or equal to the interest information threshold, or whether the interest information is among the first N of multiple interest information, where N is a positive integer. If the interest information meets the target condition, it indicates that the target account has a high level of interest in the target object at the current moment; if the interest information does not meet the target condition, it indicates that the target account has a low level of interest in the target object at the current moment.

[0166] 405. When the interest information meets the target condition, the server pushes information related to the target object to the target account.

[0167] Among them, the information related to the target object includes the live broadcast room link, video, article and other information related to the target object, which is not limited in the embodiment of the present application.

[0168] In a possible implementation, when the interest information is greater than or equal to an interest information threshold, the server pushes information related to the target object to the target account.

[0169] In this implementation, if the target account's information about the target object is greater than or equal to the interest information threshold, it indicates that the target account has a high level of interest in the target object, and when information related to the target object is pushed to the target account, the target account is more likely to accept it. If the information related to the target object is a link to the target object's live broadcast room, then the target account is more likely to enter the target object's live broadcast room. If the target object sells goods in the live broadcast room, then the target account is more likely to make purchases in the live broadcast room.

[0170] In one possible implementation, when the interest information is the first N interest information among multiple interest information, the server pushes information related to the target object to the target account, and the multiple interest information is the interest information of multiple accounts in the account set for the target object at the current moment, wherein N is determined by the target object, or is set by technical personnel according to actual conditions, and the embodiments of the present application do not limit this.

[0171] Among them, when the server determines the interest information of other accounts in the account set on the target object at the current moment, it belongs to the same inventive concept as determining the information of the target account on the target object. The implementation process is described in the above steps 401-403 and will not be repeated here.

[0172] Under this implementation, the server can sort the interest information corresponding to multiple accounts in the account set. Since the interest information can reflect the account's interest level in the target object, when the interest information of the target account is in the top N of the multiple interest information, it means that the target account has a higher level of interest in the target object in the account set, and the effect of pushing information related to the target object to the target account is better.

[0173] In one possible implementation, the server can also push information related to the target object to the accounts corresponding to the first N interest information in the multiple interest information. In the case where N is determined by the target object, the target object can determine the number of N by paying a certain amount of virtual currency to the server, and N is positively correlated with the amount of virtual currency paid. For example, celebrity A wants to start a live broadcast to sell goods, then celebrity A can pay a certain amount of virtual currency on the server to purchase traffic. The server determines N based on the virtual currency paid, and pushes the link to the live broadcast room of celebrity A to the accounts corresponding to the first N interest information in the multiple interest information.

[0174] Through the technical solution provided by the embodiment of the present application, when making information recommendations, the first interest feature of the target account for the target object in the target time period is determined. This first interest feature combines the browsing information of the target object by multiple accounts in the account set, and the browsing information of the target object by accounts other than the target account in the multiple accounts can be regarded as a baseline. The first interest feature determined based on this baseline can more accurately reflect the target account's interest in the target object. When determining the interest information, the time difference and the interest decay parameter are combined, that is, the influence of time on the interest level is taken into account, and the obtained interest information can more accurately reflect the target account's interest level in the target object at the current moment. Pushing information related to the target object to accounts whose interest information meets the target conditions also improves the accuracy of information push.

[0175] Figure 8 FIG. 1 is a block diagram of an information push device according to an exemplary embodiment. Figure 8 The device includes a first interest feature determination unit 801, an interest information determination unit 802 and a push unit 803.

[0176] The first interest feature determination unit 801 is configured to determine the first interest feature of the target account for the target object within the target time period based on browsing information of multiple accounts in the account set for the target object within the target time period, and the first interest feature is used to represent the degree of interest of the target account in the target object, and the target account belongs to the account set.

[0177] The interest information determination unit 802 is configured to determine the interest information of the target account in the target object based on the time difference between the target time period and the current moment, the interest decay parameter and the first interest feature. The interest information is used to indicate the degree of interest of the target account in the target object at the current moment, and the interest decay parameter is used to indicate the degree of decay of the interest degree over time.

[0178] The push unit 803 is configured to push information related to the target object to the target account when the interest information meets the target condition.

[0179] In one possible implementation, the first interest feature determination unit 801 is configured to execute acquisition of a first browsing count and a second browsing count corresponding to the target account within the target time period, wherein the first browsing count is the number of browsing counts of the target account for the target object, and the second browsing count is the number of browsing counts of the target account for objects in the object set to which the target object belongs, the object set including multiple objects, and the target object belongs to the object set. The first number of accounts and the second number of accounts corresponding to the target time period are acquired, wherein the first number of accounts is the number of accounts in the account set that have browsed any object in the object set within the target time period, and the second number of accounts is the number of accounts in the account set that have browsed the target object within the target time period. The first interest feature is determined based on the first browsing count, the second browsing count, the first number of accounts, and the second number of accounts.

[0180] In one possible implementation, the first interest feature determination unit 801 is configured to determine the quotient of the first number of views and the second number of views as the first viewing frequency of the target object by the target account; determine the quotient of the number of first accounts and the number of second accounts as the second viewing frequency of the target object by the multiple accounts; and determine the product of the first viewing frequency and the second viewing frequency as the first interest feature.

[0181] In one possible implementation, the interest information determination unit 802 is configured to determine the time decay value of the target account for the target object based on the time difference and the interest decay parameter, where the time decay value is negatively correlated with the time difference. The product of the first interest feature and the time decay value is determined as a second interest feature, and the second interest feature is used to represent the degree of interest of the target account in the target object at the current moment. The second interest feature is normalized to obtain the interest information. The second interest feature is normalized to obtain the interest value.

[0182] In one possible implementation, there are multiple target time periods, and the first interest feature determination unit 801 is configured to determine multiple first interest features of the target account for the target object in the multiple time periods based on browsing information of multiple accounts in the account set for the target object in the multiple time periods. The interest information determination unit 802 is configured to determine the target account's interest information for the target object based on multiple time differences between the multiple target time periods and the current moment, the interest decay parameter, and the multiple first interest features.

[0183] In one possible implementation, the first interest feature determining unit 801 is configured to determine an initial second interest feature corresponding to each target time period based on the time difference between each target time period and the current moment, the interest decay parameter, and the first interest feature corresponding to each target time period. Based on the initial second interest feature corresponding to each target time period, information about the target account's interest in the target object is obtained.

[0184] In one possible implementation, the first interest feature determining unit 801 is configured to fuse the initial second interest features corresponding to each target time period to obtain the second interest feature of the target account for the target object, and normalize the second interest feature to obtain the interest information of the target account for the target object.

[0185] In a possible implementation, the push unit 803 is configured to perform any of the following:

[0186] When the interest information is greater than or equal to the interest information threshold, information related to the target object is pushed to the target account.

[0187] In the case where the interest information is the first N interest information among multiple interest information, information related to the target object is pushed to the target account, N is a positive integer, and the multiple interest information is the interest information of multiple accounts in the account set on the target object at the current moment.

[0188] In a possible implementation, the time difference between the target time period and the current time is any one of the following:

[0189] The difference between the earliest time in the target time period and the current time.

[0190] The difference between the midpoint of the target time period and the current time.

[0191] Regarding the apparatus in the above embodiment, the specific manner in which each unit performs operations has been described in detail in the embodiment of the method, and will not be elaborated on here.

[0192] Through the technical solution provided by the embodiment of the present application, when making information recommendations, the first interest feature of the target account for the target object in the target time period is determined. This first interest feature combines the browsing information of the target object by multiple accounts in the account set, and the browsing information of the target object by accounts other than the target account in the multiple accounts can be regarded as a baseline. The first interest feature determined based on this baseline can more accurately reflect the target account's interest in the target object. When determining the interest information, the time difference and the interest decay parameter are combined, that is, the influence of time on the interest level is taken into account, and the obtained interest information can more accurately reflect the target account's interest level in the target object at the current moment. Pushing information related to the target object to accounts whose interest information meets the target conditions also improves the accuracy of information push.

[0193] In the embodiment of the present application, the electronic device can be implemented as a server. The structure of the server is described below:

[0194] Figure 9 1 is a block diagram of a server 900 according to an exemplary embodiment. The server 900 may vary significantly due to different configurations or performance, and may include one or more processors (Central Processing Units, CPUs) 901 and one or more memories 902. The memory 902 stores at least one instruction, which is loaded and executed by the processor 901 to implement the above-mentioned information push method.

[0195] In an exemplary embodiment, a computer-readable storage medium including instructions is further provided, such as a memory including instructions. The instructions can be executed by the processor 901 of the server 900 to implement the above-mentioned information push method. Optionally, the storage medium can be a non-transitory storage medium, such as a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0196] In an exemplary embodiment, a computer program product is further provided, including a computer program, which can be executed by a processor of an electronic device to implement the above-mentioned information push method.

[0197] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0198] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

[0199] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0200] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. An information push method, characterized in that: include: Obtaining a first view count and a second view count corresponding to a target account in each of multiple target time periods, where the target account belongs to an account set, the first view count being the number of views of the target object by the target account, and the second view count being the number of views of an object in the object set to which the target object belongs; Obtaining a first number of accounts and a second number of accounts corresponding to each target time period, wherein the first number of accounts is the number of accounts in the account set that have browsed any object in the object set during the target time period, and the second number of accounts is the number of accounts in the account set that have browsed the target object during the target time period; Determining a first interest feature corresponding to each target time period based on the first number of views, the second number of views, the number of the first accounts, and the number of the second accounts corresponding to each target time period, where the first interest feature is used to indicate the degree of interest of the target account in the target object; Determining an initial second interest feature corresponding to each target time period based on a time difference between each target time period and the current moment, an interest decay parameter, and a first interest feature corresponding to each target time period, wherein the interest decay parameter is used to indicate a degree of decay of the interest level over time; fusing the initial second interest features corresponding to the target time periods to obtain a second interest feature of the target account for the target object, where the second interest feature is used to indicate the degree of interest of the target account in the target object at the current moment; Normalizing the second interest feature to obtain interest information of the target account in the target object, where the interest information is used to indicate the degree of interest of the target account in the target object at the current moment; When the interest information meets the target condition, information related to the target object is pushed to the target account.

2. The information push method according to claim 1, characterized in that: The determining the first interest feature based on the first browsing count, the second browsing count, the number of the first accounts, and the number of the second accounts includes: determining the quotient of the first browsing number and the second browsing number as a first browsing frequency of the target object by the target account; determining the quotient of the first number of accounts and the second number of accounts as a second browsing frequency of the target object by the multiple accounts; The product of the first browsing frequency and the second browsing frequency is determined as the first interest feature.

3. The information push method according to claim 1, wherein: The method further comprises: Determining a time decay value of the target account to the target object based on the time difference and the interest decay parameter, wherein the time decay value is negatively correlated with the time difference; The product of the first feature of interest and the time decay value is determined as the second feature of interest.

4. The information push method according to claim 1, wherein: When the interest information meets the target condition, pushing information related to the target object to the target account includes any of the following: When the interest information is greater than or equal to the interest information threshold, pushing information related to the target object to the target account; In the case where the interest information is the first N interest information among multiple interest information, information related to the target object is pushed to the target account, N is a positive integer, and the multiple interest information is the interest information of multiple accounts in the account set on the target object at the current moment.

5. The information push method according to any one of claims 1 to 4, characterized in that: The time difference between the target time period and the current time is any of the following: The difference between the earliest time in the target time period and the current time; The difference between the midpoint of the target time period and the current time.

6. An information push device, characterized in that: include: a first interest feature determination unit configured to obtain a first browsing count and a second browsing count corresponding to a target account in each of a plurality of target time periods, wherein the target account belongs to an account set, the first browsing count being the number of browsing counts of a target object by the target account, and the second browsing count being the number of browsing counts of an object in the object set to which the target object belongs; Obtaining a first number of accounts and a second number of accounts corresponding to each target time period, wherein the first number of accounts is the number of accounts in the account set that have browsed any object in the object set during the target time period, and the second number of accounts is the number of accounts in the account set that have browsed the target object during the target time period; Determining a first interest feature corresponding to each target time period based on the first number of views, the second number of views, the number of the first accounts, and the number of the second accounts corresponding to each target time period, where the first interest feature is used to indicate the degree of interest of the target account in the target object; The interest information determination unit is configured to determine an initial second interest feature corresponding to each target time period based on a time difference between each target time period and the current moment, an interest decay parameter, and a first interest feature corresponding to each target time period, wherein the interest decay parameter is used to indicate a degree of decay of the interest level over time; fuse the initial second interest features corresponding to each target time period to obtain a second interest feature of the target account for the target object, wherein the second interest feature is used to indicate a degree of interest of the target account in the target object at the current moment; and normalize the second interest feature to obtain interest information of the target account for the target object, wherein the interest decay parameter is used to indicate a degree of decay of the interest level over time. The push unit is configured to push information related to the target object to the target account when the interest information meets the target condition.

7. The information push device according to claim 6, characterized in that: The first interest feature determination unit is configured to determine a quotient of the first browsing count and the second browsing count as a first browsing frequency of the target object by the target account; The quotient of the first number of accounts and the second number of accounts is determined as the second browsing frequency of the target object by multiple accounts; and the product of the first browsing frequency and the second browsing frequency is determined as the first interest feature.

8. The information push device according to claim 6, characterized in that: The interest information determination unit is further configured to determine the time decay value of the target account to the target object based on the time difference and the interest decay parameter, and the time decay value is negatively correlated with the time difference; and determine the product of the first interest feature and the time decay value as the second interest feature.

9. The information push device according to claim 8, characterized in that: The pushing unit is configured to perform any of the following: When the interest information is greater than or equal to the interest information threshold, pushing information related to the target object to the target account; In the case where the interest information is the first N interest information among multiple interest information, information related to the target object is pushed to the target account, N is a positive integer, and the multiple interest information is the interest information of multiple accounts in the account set on the target object at the current moment.

10. The information push device according to any one of claims 6 to 9, characterized in that: The time difference between the target time period and the current time is any of the following: The difference between the earliest time in the target time period and the current time; The difference between the midpoint of the target time period and the current time.

11. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the information push method according to any one of claims 1 to 5.

12. A computer-readable storage medium, characterized in that When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the information pushing method according to any one of claims 1 to 5.

13. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the information pushing method according to any one of claims 1 to 5 is implemented.

Images