CN116150488A

CN116150488A - Method, apparatus, device, medium and program product for presenting recommended content

Info

Publication number: CN116150488A
Application number: CN202310129545.1A
Authority: CN
Inventors: 董重里; 徐卓鑫; 刘瑞珺; 李端
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2023-01-31
Filing date: 2023-01-31
Publication date: 2023-05-23

Abstract

Embodiments of the present application provide methods, apparatus, electronic devices, computer storage media, and program products for presenting recommended content. The method includes receiving an indication of first recommended content of an application presented on an application card of a user device, and determining a first preference including content preferences and formal preferences for content presentation in response to receiving the indication. The first preference is associated with a current environment in which the user device is located and is different from a second preference for presenting second recommended content within an interface within the application when the indication is received. The method also includes causing the first recommended content to be presented on the application card in a manner consistent with the first preference. According to the method, different recommendation logics are adopted for the application card and the application, so that the recommendation content and the presentation form of the application card are more fit with the instant interests of the user in a specific environment, the usefulness of the application card as a shortcut inlet of the application is enhanced, and the possibility that the user jumps into the application through the card is improved.

Description

Method, apparatus, device, medium and program product for presenting recommended content

Technical Field

The present application relates to the field of information technology, and more particularly, to a method, apparatus, electronic device, computer readable storage medium, and program product for presenting recommended content.

Background

With the increasing development of network communication technology, news information can be delivered to user devices through a network at near real-time speeds. News information is commonly acquired by information-based Applications (APP) such as cell phones. Since each user is able to receive limited information and the interests of different users are different, the news that each user expects to view is different. Accordingly, content providers of applications often recommend news to users on a recommendation page (e.g., top page) within the application based on the user's browsing records and news hotness, etc.

In addition, more and more APP provides gadgets in the form of cards that can be presented on the desktop and negative one-screen of the device, etc. These application cards may also present the recommended content of the APP to the user. Thus, the user can quickly preview the recommended news information through the card of the application without opening the APP interface. And deciding to enter corresponding APP viewing details according to the previews of the cards.

Disclosure of Invention

The embodiment of the application provides a scheme for presenting recommended content.

In a first aspect, there is provided a method of presenting recommended content, the method comprising: receiving an indication of presenting first recommended content of an application on an application card of a user device; in response to receiving the indication, determining a first preference comprising content preferences and formal preferences of content presentation, the first preference being associated with a current environment in which the user device is located, and the first preference being different from a second preference for presenting second recommended content within an application interface of the application when the indication is received; and causing presentation of first recommended content on the application card in a manner consistent with the first preference, wherein the first recommended content is determined based on the first preference, the first recommended content comprising content elements of a set of content items that are presented on the application card as representations of the set of content items. Therefore, the recommended content and the presentation form of the application card can be more fit with the instant interests of the user in the specific environment, so that the usefulness of the application card as a shortcut entry of the application is enhanced, and the possibility that the user jumps into the application through the card is improved.

In some embodiments of the first aspect, the method further comprises: if the recommended page in the application interface is detected to be opened when the indication is received, presenting second recommended content in the recommended page in a manner consistent with the second preference. Therefore, the recommended contents of the application card and the recommended page in the application have different emphasis points, and the requirements of users when accessing the application in different modes are adapted.

In some embodiments of the first aspect, the formal preferences include preferences for one or more of: recommending the media type of the content; an element type representing the content element being presented as recommended content; a layout style in which recommended content is presented on an application card; and a visual style in which the recommended content is presented on the application card. Thus, the first recommended content may be determined and presented in a form consistent with these aspects.

In some embodiments of the first aspect, determining the first recommended content based on the first preference comprises: determining a media type of the first recommended content based on the formal preference; and determining a set of content items associated with the content preferences and having the media type as the set of content items. Therefore, the type and the topic of the recommended content can be enabled to accord with the instant interest of the user.

In some embodiments of the first aspect, determining the first recommended content based on the first preference further comprises: determining an element type of the content element based on the form preference; and extracting content elements associated with the content preferences and having element types from the content items in the determined set of content items based on the content preferences and the element types. Therefore, the expression mode of the recommended content item presented to the user can be enabled to accord with the instant interest of the user.

In some embodiments of the first aspect, further comprising: determining at least one of a layout style and a visual style of the content element in the application card based on the form preference; and typesetting the content elements according to at least one of the determined layout style and visual style; and causing the first recommended content to be presented on the application card in a manner consistent with the first preference includes causing the typeset content elements to be presented on the application card. Therefore, the presentation mode of the card content can be enabled to accord with the instant interest of the user.

In some embodiments of the first aspect, determining the first preference comprises: determining a user with whom the application is currently associated on the user device; and determining a first preference for a feature of the current environment based on a correlation between a historical behavioral feature of the user accessing the content and a historical environmental feature. Thus, the determined preferences can be made to conform to the historical interests of a particular user in a particular environment, making the recommendation targeted.

In some embodiments of the first aspect, determining the first preference for the current environment comprises: preferences are determined for the current environment based on the correlations and profile information of the user. Thus, the determined instant preference may take into account the personal characteristics of the user and the explicitly set interests, making the recommended content based thereon more targeted.

In some embodiments of the first aspect, the method further comprises determining the correlation by: acquiring a set of records about a user, wherein the records in the set of records comprise behavior data describing characteristics of respective behaviors of the user accessing content and environment data describing characteristics of an environment in which a user device for performing the respective behaviors is located when the respective behaviors are performed; and determining a correlation by extracting historical correlations between the behavioral data and the environmental data in the set of records. Thus, historical preferences of the user associated with the environmental scenario may be obtained, providing an effective basis for determining the instant interests of the user.

In some embodiments of the first aspect, the set of records is a first set of records and is acquired during a first period of time, and the method further comprises: acquiring a second set of records for the user during a second time period subsequent to the first time period; and updating the determined correlation based on the historical correlation between the behavioral data and the environmental data in the second set of records. Thus, the logic for recommending instant preferences may be made non-obsolete by the user's preferences changing.

In a second aspect, there is provided an apparatus for presenting recommended content, the apparatus comprising: a receiving module configured to receive an indication of first recommended content of an application presented on an application card of a user device; a determining module configured to determine, in response to receiving the indication, a first preference comprising a content preference and a formal preference for presentation of the content, the first preference being associated with a current environment in which the user device is located, and the first preference being different from a second preference for presentation of second recommended content within an application interface of the application when the indication is received; and a presentation module configured to cause presentation of first recommended content on the application card in a manner consistent with the first preference, wherein the first recommended content is determined based on the first preference, the first recommended content comprising content elements of a set of content items that are presented on the application card as representations of the set of content items.

In some embodiments of the second aspect, the apparatus further comprises: and a second presentation module configured to cause presentation of second recommended content within the recommended page in a manner consistent with the second preference if the recommended page within the application interface is detected to be opened upon receipt of the indication.

In some embodiments of the second aspect, wherein the formal preferences include preferences for one or more of: recommending the media type of the content; an element type representing the content element being presented as recommended content; a layout style in which recommended content is presented on an application card; and a visual style in which the recommended content is presented on the application card.

In some embodiments of the second aspect, the presentation module comprises: a media type determination module configured to determine a media type of the first recommended content based on the formal preference; and a content item determination module configured to determine a set of content items associated with the content preference and having a media type as a set of content items.

In some embodiments of the second aspect, the presentation module further comprises: an element type determination module configured to determine an element type of the content element based on the formal preference; and a content element determination module configured to extract, based on the content preferences and the element types, content elements associated with the content preferences and having the element types from the content items in the determined set of content items.

In some embodiments of the second aspect, the presentation module comprises: a composition determination module configured to determine at least one of a layout style and a visual style of the content elements in the application card based on the form preference; a typesetting module configured to typeset the content elements according to at least one of the determined layout style and visual style; and the presentation module includes a card loading module configured to cause the typeset content elements to be presented on the card.

In some embodiments of the second aspect, the determining module comprises: a user determination module configured to determine a user with whom the application is currently associated on the user device; and a preference determination module configured to determine a first preference for a feature of the current environment based on a correlation between a historical behavioral characteristic of the user accessing the content and a historical environmental characteristic.

In some embodiments of the second aspect, the preference determination module comprises: the comprehensive determination module determines preferences for the current environment based on the correlation and profile information of the user.

In some embodiments of the second aspect, the apparatus further comprises: an acquisition module configured to acquire a set of records about a user, wherein the records in the set of records include behavior data describing characteristics of respective behaviors of the user accessing the content and environment data describing characteristics of an environment in which a user device for performing the respective behaviors is located when the respective behaviors are performed; and a relationship determination module configured to determine a correlation by extracting historical correlations between the behavioral data and the environmental data in the set of records.

In some embodiments of the second aspect, the set of records is a first set of records and is acquired during a first period of time, and the apparatus further comprises: a second acquisition module configured to acquire a second set of records of the user for a second period of time after the first period of time; and a relationship updating module configured to update the determined correlation based on the historical correlation between the behavior data and the environmental data in the second set of records.

In a third aspect, an electronic device is provided. The electronic device comprises a processor and a memory having stored thereon computer instructions which, when executed by the processor, cause the electronic device to perform the actions of the method according to the first aspect or any embodiment thereof described above.

In a fourth aspect, a computer-readable storage medium is provided. The computer-readable storage medium has stored thereon computer-executable instructions which, when executed by a processor, implement operations according to the method of the first aspect described above or any embodiment thereof.

In a fifth aspect, a computer program or computer program product is provided. The computer program or computer program product is tangibly stored on a computer-readable medium and comprises computer-executable instructions which, when executed, implement operations in accordance with the method in the first aspect or any embodiment thereof described above.

Drawings

The above and other features, advantages, and aspects of embodiments of the present application will become more apparent by reference to the following detailed description when taken in conjunction with the accompanying drawings. In the drawings, wherein like or similar reference numerals denote like or similar elements, in which:

FIG. 1 illustrates a schematic diagram of an example environment in which various embodiments of the present disclosure may be implemented;

FIG. 2 illustrates a flowchart of an example method of presenting recommended content, according to some embodiments of the disclosure;

FIG. 3 illustrates a non-limiting example of a card frame presenting a particular media type as determined based on media preferences in accordance with some embodiments of the present disclosure;

FIG. 4 illustrates a schematic diagram of a non-limiting example in which an example card frame presenting a particular type of element is shown as determined based on element type preferences, according to some embodiments of the present disclosure;

FIG. 5 shows a schematic diagram of a non-limiting example in which different elements are determined for different users of the same content item for presentation on an application card, according to some embodiments of the present disclosure;

FIG. 6 illustrates a schematic diagram of a non-limiting example in which example card layout styles determined based on layout style preferences are shown, according to some embodiments of the present disclosure;

FIG. 7 illustrates a schematic diagram of data information and its relationships involved in an example process of presenting recommended content, according to some embodiments of the present disclosure;

FIG. 8 illustrates an interface schematic of a non-limiting example according to some embodiments of the present disclosure;

FIG. 9 illustrates a schematic diagram of an example of training and use of a preference model according to some embodiments of the present disclosure;

FIG. 10 illustrates a schematic block diagram of an apparatus for presenting recommended content according to some embodiments of the disclosure; and

FIG. 11 shows a schematic block diagram of an example device that may be used to implement embodiments of the present disclosure.

Detailed Description

Embodiments of the present application will be described in more detail below with reference to the accompanying drawings. While certain embodiments of the present application are shown in the drawings, it is to be understood that the present application may be embodied in various forms and should not be construed as limited to the embodiments set forth herein, but rather are provided to provide a more thorough and complete understanding of the present application. It should be understood that the drawings and examples of the present application are for illustrative purposes only and are not intended to limit the scope of the present application.

In the description of the embodiments of the present application, the term "comprising" and its similar terms should be understood as open-ended, i.e. "including, but not limited to. The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first," "second," and the like, may refer to different or the same object. Other explicit and implicit definitions are also possible below.

In conventional approaches, when recommended content is presented using, for example, an application card (also referred to below as a card) on the negative screen of a mobile device, the determination and presentation of the recommended content still continues with the information recommendation logic within the application. Such recommendation logic does not take into account the needs of the user in different scenarios. For example, a user may be more listening to audio news while driving, and may be more inclined to watch video news while at home. For example, a user may prefer to browse the time news in the early morning and the entertainment information in the evening when leisure. In addition, when the application card interface is refreshed, only recommended content is refreshed, the form change of the card interface is almost negligible, and different content presentation modes are not provided for the user based on visual preference of the environment in which the user is located. These reasons have resulted in application cards often not being able to attract the attention of the user, relying solely on the subjective intent of the user to use the card portal, reducing the functionality of the card as a unique portal.

To solve the above-described problems, as well as other problems, the present disclosure provides a solution for presenting recommended content. The scheme determines, upon receiving an indication of recommended content of the application presented on an application card of the user device (e.g., triggered by an operation to refresh the card interface), an immediate user preference comprising content preferences and formal preferences for content presentation. The instant user preference is different from the preference used at this time to present the recommended content within the application interface of the application and is associated with the environment in which the user device is currently located.

Then, when refreshing the card interface, the scheme can determine card presentation content based on the content preference, determine the presentation form of the card presentation content based on the form preference, and then load the card presentation content onto the card in the determined presentation form, so as to enable the corresponding recommended content to be presented on the application card in a manner consistent with the instant user preference. In some embodiments, the scheme may determine a user to whom the application is currently associated (e.g., logged on) on the user device, and determine an immediate preference of the user in the current environment for the current environment based on a correlation between historical behavior of the user accessing the content and the historical environment.

The scheme of the present disclosure uses different mechanisms than recommendation logic within an application to make differentiated content recommendations for application cards and takes into account user interest changes in different scenarios in the recommendation. Therefore, the recommended content and the presentation form of the application card can be more fit with the instant interests of the user in the specific environment, so that the usefulness of the application card as a shortcut entry of the application is enhanced, and the possibility that the user jumps into the application through the card is improved.

FIG. 1 illustrates a schematic diagram of an example environment 100 in which various embodiments of the present disclosure may be implemented. As shown in fig. 1, an example environment 100 includes a user device 110, a user 120, and a computing device 130. The user device 110 may be an electronic device with an application installed. In fig. 1, the user device 110 is shown as a smart phone, but in other embodiments, the user device 110 may also be implemented as other forms of portable terminal devices, such as a tablet computer, as well as a wearable device.

The user device 110 has an operating system installed therein, which has a system program built therein, and also supports the installation of third party applications. The operating system of the user device 110 may support the independent presentation and functioning of the applications as application cards with portions of the services or functions on the desktop interface of the user device 110. As shown in FIG. 1, an application card 140 associated with a particular application (not shown), such as a news application, is displayed on the negative screen of user device 110. The application card 140 may have content for the application presented thereon. For example, the application may be a news application and the recommended news may be presented on application card 140. It should be appreciated that in some embodiments, the application card 140 may also be a gadget card that is added to other screens of the desktop interface.

The user 120 may operate the user device 110, move around with the user device 110, and so on. For example, the user 120 may log in an account associated therewith on the user device 100. For example, user device 110 may operate user device 110 to add application card 140 to the negative one of user device 110. User 120 may also preview the presentation of application card 140 and view more detailed content by clicking on application card 140 to enter the application. In some embodiments, the user 120 may also connect the user device 110 to an external device, such as an in-car stereo, an external screen, or the like.

User device 110 may communicate with computing device 130. It should be understood that embodiments of the present disclosure are not limited in the manner of communication, such as communication in a wired or wireless manner. The wired manner may include, but is not limited to, fiber optic connections, universal serial bus (Universal Serial Bus, USB) connections, etc., and the wireless manner may include, but is not limited to, mobile communication technologies including, but not limited to, cellular mobile communications, etc., wi-Fi, bluetooth (Bluetooth), point-to-Point (P2P), etc. furthermore, during movement of the user device 110, the manner in which it communicates with the computing device 130 may change, e.g., transition from cellular mobile network communications to Wi-Fi communications.

In some embodiments, the computing device 130 may be a server of a content provider. For example, computing device 130 may be an operator device of an application associated with application card 140. Computing device 130 may provide content to applications on user device 110. For example, the computing device 130 may use methods in accordance with embodiments of the present disclosure to cause recommended content to be presented on the application card 140. It should be appreciated that although shown as a single device, computing device 130 may be implemented in any other form suitable for performing the respective functions, such as a plurality of centralized devices, distributed devices, and/or deployed in the cloud, among others.

It should be understood that the architecture and functionality in the example environment 100 are described for illustrative purposes only and are not meant to suggest any limitation as to the scope of the disclosure. For example, in some embodiments, the computing device 130 may also cause respective recommended content to be presented on application cards of a plurality of other user devices according to methods of embodiments of the present disclosure. Also, other devices, systems, or components, etc., not shown, may also be present in the example environment 100.

In addition, embodiments of the present disclosure may also be applied in other environments having different structures and/or functions. For example, in some embodiments, the user device 110 may also perform methods according to embodiments of the present disclosure to determine the user's immediate preferences locally, retrieve the corresponding content from the provider of the application for typesetting for presentation on the application card 140, and so on.

FIG. 2 illustrates a flowchart of an example method 200 of presenting recommended content, according to some embodiments of the disclosure. The example method 200 may be performed, for example, by the computing device 130 shown in fig. 1. For example, the computing device 130 may perform the example method 200 to cause recommended content to be presented on the card 140. It should be appreciated that method 200 may also include additional actions not shown, the scope of the present disclosure being not limited in this respect. The method 200 is described in detail below in connection with the example environment 100 of fig. 1.

At block 210, an indication of first recommended content of an application presented on an application card of a user device is received. For example, the computing device 130 may receive an indication (such as card content initialization, card content refresh, etc.) to present the first recommended content of the application on the application card 140 of the user device 110.

In some embodiments, the indication may be triggered by a specific operation of the user 120 or in response to a signal detecting certain events. For example, all content on the negative screen, including the application card 140, may be refreshed each time the user 120 operates the user device 110 to reenter the negative screen. For example, the content of the gadget card on the other screen may also be refreshed each time a cut is made into the page on which the card is located. For example, the application card may be configured with a refresh control, and when the user 120 clicks on the refresh control, the content of the application card may be refreshed.

In response to receiving the indication, a first preference including content preferences and formal preferences for presentation of the content is determined at block 220. The first preference is associated with a current environment in which the user device is located and is different from a second preference for presenting second recommended content within an application interface of the application when the indication is received. For example, in response to receiving an indication to present recommended content on application card 140, computing device 130 may determine a first preference associated with a current environment in which user device 110 is located. At this point, if the user opens the recommendation page in a conventional manner (e.g., clicking on the application icon initiates a top page that includes recommended content), the computing device 130 presents the recommended content within the recommendation page that is different from the card content preferences, as will be detailed later.

The first preference includes a content preference and a formal preference for presentation of the content. The content preferences indicate preferences on the content, which may include topics of content sources of interest, topics of content elements presented on cards, and so forth. The form preference indicates a user's presentation form preference at the card interface, such as the media type of the content source, the card layout style, the media type of the content elements presented, how much and where on the card the text is located, etc. Thus, each time a refresh of a card page is performed, the presentation content and the presentation form of the application card 140 may be updated based on the determined first preference.

In some embodiments, computing device 130 may gather (e.g., via user device 110) environmental information, such as physical environmental information, network status information, and available media device (e.g., speakers, etc.) information, in which user device 110 is currently located. The computing device 130 may then determine the first preference based on the environmental information. For example, the computing device 130 may make predictions of the first preference using the environmental information as input to a preference model, as described below.

In some embodiments, the first preference is associated with both the current environment and the user with whom the application is currently associated on the user device 110. In other words, the first preference is a preference of the current user in the current environment. In some embodiments, computing device 130 may determine a user with whom the application is currently associated on the user device. For example, computing device 130 may determine that an account of user 120 is currently logged on an application associated with application card 140, or may determine that a general account of user 120 is logged on user device 110. The computing device 130 may then determine an immediate preference of the user 120 in the current environment as the first preference.

In some embodiments, the computing device 130 may determine the first preference for the current environmental feature based on a correlation between historical behavioral characteristics of the user 120 accessing the content and historical environmental characteristics. For example, the computing device 130 may predict the user's immediate preferences using a user preference model, which may be determined based on the user's browsing behavior data. The construction and use of the preference model will be described in more detail later with respect to fig. 8.

On this basis, in predicting the instant preference of the user 120, the computing device 130 may input the instant environment information (e.g., in the form of an environment tag) in which the user device 110 is located as a variable input to the trained preference model for the user 120. The user preference model may then output the instant preferences for the user 120, for example, in the form of a set of preference tags. The instant preference may then be used to determine recommended content presented on the application card 140. The determination of the instant preference will be described in more detail later with reference to the examples of tables 1 to 5.

In some embodiments, in addition to the user usage data, computing device 130 also considers profile information generated at the time of user registration when predicting user preferences. For example, the computing device 130 may determine its immediate preferences for the current environment for both the aforementioned correlations of the user 120 and the profile information of the user 120. In such embodiments, the impact of the user profile information may be considered in constructing the predictive model, as described later in connection with fig. 8.

At block 230, the first recommended content is caused to be presented on the application card in a manner consistent with the first preference. For example, the computing device 130 may cause the first recommended content to be presented on the application card 140 in a manner consistent with the first preference determined at block 220. The computing device 130 may determine, based on the immediate preference of the user 120 determined at block 220, a first recommended content that meets the preference, the first recommended content comprising content elements in a set of content items. These content elements should be actually presented as filler content on the application card 140, the content elements of the respective content item representing the content item. The content type, media type and filler content of the first recommended content, as well as the layout style and visual style it is presented to conform to the immediate preference of the user 120 in the current environment.

The computing device 130 may then type out the elements that are filler content in a manner that meets the immediate preference and cause them to be presented on the application card 140. For example, the computing device 130 may instruct the user device 110 to incorporate the typeset first recommended content into the display data stream of the screen and present on the card 140 when the corresponding screen or card 140 is refreshed. The determination and presentation of the first recommended content will be described in more detail later with respect to fig. 4-7.

Using the method 200, content recommendation for application cards using mechanisms that are not exactly the same as recommendation logic within the application, and taking into account the user's preference changes in different scenarios, may cause at least one of the recommended content and its presentation to change with the environment in which the user is located, and/or with the user. For example, some users' preferences for content topics may not change with the environment, so recommended content items on the card and within the application at the same time may be approximately the same. However, due to additional considerations of the environment, the presentation form of the application card may still vary from the recommended pages within the application as the environment changes. Therefore, the content recommendation of the application card can be more in line with the instant interest of the user in the specific environment, and the usefulness of the application card as a shortcut inlet of the application is enhanced.

It should be appreciated that although described for ease of illustration in the context of computing device 130 executing method 200 separate from the user device, in some embodiments, method 200 may also be performed by user device 110, e.g., user device 110 may determine the first preference locally and retrieve the corresponding content from the content pool of the content provider remotely (e.g., via a service indication) and present on the application card in accordance with the first preference.

As previously described, the first preference is different from the second preference for presenting the second recommended content within the application interface of the application upon receipt of the indication. For application cards 140 that provide shortcut content previews, this first preference is generally more consistent with the current scenario needs. In some embodiments, if the recommendation page within the application interface is detected to be open upon receipt of the indication, the computing device 130 causes the second recommendation content to be presented within the recommendation page in a manner consistent with the second preference.

It should be understood that "upon receipt of an indication" herein is a reasonably short time (such as seconds, tens of seconds, or minutes) before and after receipt of the indication, and that the environment in which the user device 110 is located is unchanged. In the preferred embodiment for a particular user, the user associated with the application is not changed during the period of time that meets the "when indication was received". Further, it should be appreciated that the content presented by the card in conventional approaches may be a subset of the content displayed within the application (e.g., intercepting the display of the foremost portion) due to the size limitations of the card. Such subset display is still based on the same logic of recommendations, as opposed to differentiated content and/or forms of recommendations in the present disclosure.

The second recommended content may be determined and presented based on any suitable recommendation logic. For example, the computing device 130 may be more interested in the user's comprehensive or long-term interests in all environments when determining the second recommended content, and/or may be based on the user's layout settings when typesetting, etc. The present disclosure is not limited by the particular manner in which recommended content is presented within an application interface. Such an embodiment can adapt to different needs of a user when accessing an application in different ways.

An example method of determining an instant preference of a user (e.g., a first preference in method 200) that may then be used to determine a content presentation of an application card according to some embodiments of the present disclosure will now be described in more detail. The acts of these methods may be performed, for example, by computing device 130. For purposes of illustration, the following will be described in connection with the example environment 100 of FIG. 1.

Upon collecting current environmental information in which the user device 110 is located, the computing device 130 may input the information into a preference model, which may be built (e.g., trained) based on historical behavior data of the user accessing the content, to predict the instant user preferences. For example, the computing device 130 may convert the relevant information recorded by the user device into the form of an environmental label and input the environmental label into the preference model of the user 120 to obtain the current instant preference of the user 120. The preference model may output the instant preference, for example, in the form of an instant preference tag.

In some embodiments, the environment information may include scene information, media device information, and network status information. Each type of information may include a plurality of attributes. Thus, in the input of preference prediction in the form of a tag, each type of information may be converted into a plurality of tags of the corresponding type, hereinafter exemplified in the form of a tag.

The scene tag indicates at least one attribute (or feature) of the environment in which the user is currently located. For example, the scene tag may include attributes within an automobile, on a subway, on a high-speed rail, at home, at a company, outdoors, at a consumer location, etc., for example, the scene tag may also include a time attribute such as day or night. The network status tag indicates the status of the current network. For example, the network status tag may include a network type to which the user equipment 110 is connected, a network speed of the network, a stability of the network, and the like.

The media device tag indicates what device the content is to reach the user through, e.g., whether the user device 110 is connected to an external media device and the type of media device connected, etc. For example, if the user device 110 is connected to bluetooth in vehicle, the user 120 is instructed to contact the content through car audio. For example, if the user device 110 is connected to a speaker or headphones, the user 120 is instructed to obtain content through the speaker or headphones. For another example, if the user device 110 is connected to an external screen, it indicates that the user is contacting content through other projection devices.

Some of these tags may be obtained directly from the device record or sensed information data. For example, based on the geographic location sensed by the user device 110 carried by the user 120 and the map data indicating that it is at a shopping mall, it may be determined that a scene tag is included at the consumer location. Other tags may be obtained in conjunction with user settings information on the user device 110. For example, the user 120 has set a location of a home (or company) on the device, and in conjunction with the current geographic location of the user device 110 being the same as the set home (company), it may be determined that the scene tag is included at home (or at company).

The computing device 130 may then cause the preference model to output predicted immediate preferences based on the entered tags. The instant preferences include content preferences and form preferences. Content preferences may include, for example, preferences for topics and levels of detail, etc. The formal preferences may include preferences for media type, element type of actual presentation, typesetting layout, visual style, and so forth. Similar to the input, the preference model may output the instant preference, for example, in the form of a set of preference tags.

In some embodiments, the input-output relationship of the preference model conforms to the relationship between corresponding data in the user history data used to train the model. The user history data may include behavioral characteristics of the user's access content and characteristics of the environment in which the behavior occurred (e.g., converted to a tagged form). Various features of the environment may have an impact on various behavioral trends of the user in browsing the content. During training, the preference model may learn correlations between these attributes and the behavioral tendencies of the user. The preference model may be trained using any suitable manner, and embodiments of the present disclosure are not limited to a particular training manner of the preference model.

The media type preference indicates a user's preference for the media type of the content item presented on the application card. The media types may include, for example, video, audio, pictures, text, and the like. Preferences for one or more media types may be presented simultaneously on the card. For example, media type preferences may be affected by at least one characteristic of real-time scenarios, network status, and available media device types, for example. The impact (in other words, the correlation between the relevant environmental features and preferences) may also be determined from historical data of the user (e.g., through training of a predictive model), and may be different for different users.

Table 1 shows a non-limiting simplified example of a correlation between individual environmental characteristics and media type preferences that can be learned from user data. As can be seen from table 1, the user prefers to browse video type content at home and text type at company. Similarly, the user prefers to browse video when the network speed is fast and to browse text when the network speed is slow, and the lack of network speed from the user data has a preferential impact on the user's browsing audio.

Table 1: example of the impact of a single environmental feature on media type preferences

Media type	Video frequency	Text with a character pattern	Audio frequency	Picture picture
					Real-time location	At home	At a company	In a motor vehicle	On the subway
Medium apparatus	Connected to the screen-carrying apparatus	Medium-free device	Is connected to the sound box	Medium-free device
					Network speed	Quick-acting toy	Slow down	/	Faster
Network type	wifi	Mobile network	/	wifi
					Network stability	High height	Low and low	/	In (a)

In some embodiments, the immediate preference may be determined by a plurality of real-time environmental features, and the plurality of real-time environmental features may have an adverse effect on the user's preferences. Thus, the preference model may determine the immediate preference based on the combined impact of the plurality of environmental features. Table 2 shows a simplified example of the correlation between the combination of scene and network speed and media type preferences that can be learned from user data. As can be seen from table 2, when the user is at home and the network speed is fast, video type content is preferred and text is preferred when the network speed is slow. On the other hand, the user prefers text content at the company regardless of the internet speed.

Table 2: examples of the impact of scenario and network speed on media type

	At home	At a company	In a motor vehicle	On the subway
					Network speed is high	Video frequency	Text with a character pattern	Audio frequency	Picture picture
Network speed is slow	Text with a character pattern	Text with a character pattern	Text with a character pattern	Text with a character pattern

Element type preference indications refer to user preferences for the type of content element actually populated on the card. The element types may include, for example, a title, a summary extracted from a body text, a popular comment, a thumbnail or detail drawing extracted from a body text, a frame selected from video content, and the like. For example, content type preferences may be affected by real-time scenes such as real-time location and/or time. These effects may also be determined from historical data of the user and may be different for different users.

In one example, the element type preference may indicate the following trend: the user may carefully choose the content to view at home, preferring to choose the content by viewing "title + profile"; users, when in company, are primarily quick preview information, tend to only view titles; and users tend to read extra leisure on subways, which may be of interest to popular reviews in addition to the "title + profile". Furthermore, different time periods may also have an impact on the user's element type preferences. Table 3 shows a simplified example of the impact of scene and time on content type preferences. As can be seen from table 3, the user prefers to obtain more information of the same content item by "title + profile" at night of the company.

Table 3: examples of the impact of scene and time on element types

	At home	At a company	On the subway
				Morning of course	Title + brief description	Title of the book	Title + brief introduction + hot comments
Afternoon	Title + brief description	Title of the book	Title + brief introduction + hot comments
				At night	Title + brief introduction + hot comments	Title + brief description	Title + brief introduction + hot comments

The layout style preference indicates style preferences of the user on how the content elements are laid out on the card. The layout style preference is for the entire card deck as compared to media type preferences and content type preferences for a single content item being for a single content source. The typesetting style may include, for example, a ratio of graphics to text (e.g., multiple graphics to few text, or fewer graphics to many text, etc.), a position of graphics to text (e.g., left graphics to right, upper graphics to context, etc., text placed in the graphics, or text placed outside the graphics, etc.), and a typesetting model (e.g., tabular, content card), etc. One or more graphic types/typesetting types can be presented on the card at the same time.

For example, layout style preferences may be affected by environmental characteristics, which may be affected by real-time environment as well as network status. These effects may also be determined from historical data of the user and may be different for different users. Table 4 shows a simplified example of the impact of scene and network speed on layout style preferences. As can be seen from table 4, when the network speed is fast, the user prefers a more graphic and less text layout at home and a less graphic and more text layout at the company.

Table 4: examples of the impact of scene and network speed on layout style preferences

	At home	At a company	On the subway
				Network speed is high	Multiple pictures and text	Less-picture multi-text	Multiple pictures and text
Network speed is slow	Less-picture multi-text	Pure text	Less-picture multi-text

The visual style preferences indicate visual preferences of the user for the color and/or font, etc. of the content presented on the application card. The visual styles may include, for example, background color hue, font type, font size, line spacing, paragraph spacing, and the like. For example, visual style preferences may be affected by real-time scenes such as location and/or time. These effects may also be determined from historical data of the user and may be different for different users.

In some embodiments, the visual style preferences may be represented as a set of attributes. In other embodiments, the visual style preferences may be represented as visual style types (e.g., business, science fiction, etc.) that include a variety of attributes. Table 5 shows a simplified example of the impact of scene and network speed on visual style preferences, expressed as visual style types, that can be learned from user data. As can be seen from table 5, the user prefers a "science fiction" style during the day and a "mush style" during the evening when at home.

Table 5: examples of the impact of scene and network speed on visual style preferences

	At home	At a company	On the subway
				Daytime	Science fiction	Business business	Science and technology
Night time	Artless spread	Business business	Science fiction

The content preferences may indicate topics of interest to the user, such as football, specific clubs, automobiles, and specific brands of automobiles, among others. These effects may also be determined from historical data of the user and may be different for different users. For example, the determined content preferences may indicate that the user has no significant differences in what the user prefers in different environments. In this way, the predictive model outputs the same instant content interests based on any environment. In another example, the determined content preferences may indicate that the user's content preferences change as the environment changes, e.g., morning more preferred time news, evening more preferred entertainment news, etc. In this way, the predictive model may determine respective real-time preferences based on the respective current.

In some embodiments, the user's profile data, such as interests and age settings that they fill in when registering or updating profiles, may also be considered in determining the instant preferences. In such an embodiment, the influence of profile data may be further incorporated in the correlation model from historical data such that these settings of the user are also taken into account in the real-time input of the model predictions.

It should be appreciated that the examples described above are shown for illustration only, and that in particular implementations, the correlation of environmental features to preferences may be determined based on historical behavioral data for a target customer, and may have a more complex relationship between the two. For example, the correlation may be represented as a feature classification model such that the predictive model may output a set of instant preference tags based on the combined effect of a set of multiple environmental feature inputs.

After determining the instant preference, recommended content for the card may be presented on the application card in a manner consistent with the preference. An example method of determining and presenting recommended content for an application card according to an embodiment of the present disclosure will now be described in more detail with reference to fig. 4-7. The acts of these methods may be performed, for example, by computing device 130. For purposes of illustration, the following will be described in connection with the example environment 100 of FIG. 1.

To present the recommended content on the application card 140, the computing device 130 first performs extraction of the recommended content (e.g., the first recommended content in the method 200) based on the determined instant preference. The extraction of recommended content includes the extraction of content items to be recommended (e.g., a particular set of news) and the extraction of filler content elements. As previously mentioned, the content element is content that is specifically presented on the card.

In some embodiments, the computing device 130 may determine the media type of the first recommended content based on the formal preference of the determined immediate preferences. For example, the computing device 130 may treat some (e.g., most preferred) or all of the indicated types in the media type preferences as media types of the first recommended content. The computing device 130 may then select a set of content items (hereinafter or content sources, e.g., one or more news) that are determined to be associated with the content preference and that have the media type from the content pool of the respective application as content items to be recommended.

FIG. 3 illustrates a schematic diagram of a non-limiting example 300 showing an example card frame presenting a particular media type as determined based on media preferences, according to some embodiments of the present disclosure. Where card frame 310 presents audio content, card frame 320 presents video content, and card frame 330 presents picture content. In the population of specific content items, each card presents a Title (Title) of the corresponding content item, and card frame 320 and card frame 330 present representative frames of video and selected pictures in the picture content item, respectively. It should be appreciated that although each card in FIG. 3 presents only one media type, in a particular application, based on the instant form preference, the card may be determined to present a combination of media types, e.g., to present "audio+text", "picture+text", "video+text" type content.

In one non-limiting example, the computing device 130 may input the determined "at home, faster network speed, in network stability, not connected to an external media device" current environmental tags into the preference model. In this example, the formal preferences of the preference model output include the media types currently preferred by the user at this time being both pictorial and text, and being dominated by pictorial preferences. Further, the user's content preference tags (e.g., tags based on user-set interests and/or preferences learned by the predictive model) are determined to include "world cup" and "player A". Thus, in screening content sources, computing device 130 may screen news related to player A in world cup events from the pictorial news of the content pool and the literal news with pictures.

In some embodiments, after determining the content item of the first recommended content, the computing device 130 may determine an element type of the content element based on the formal preference. These content elements are recommended content that will actually be filled in on the card. Based on the content preferences and the determined element type, the computing device 130 may extract content elements associated with the content preferences and having the element type from content items in the set of content items determined as described above.

In some embodiments, computing device 130 may pre-extract its filler content for each element type for the content source without having to re-extract the filler content for each content type after determining element type preferences. Thus, the extracted elements of the corresponding types can be directly obtained for use after the element type preference is determined, so that the efficiency of real-time filling is improved. For example, the computing device 130 may extract a title of the content item as an element of the title type, extract a comment with the highest praise/reply number from the comments of the content item as an element of the popular comment type, and so on.

FIG. 4 illustrates a non-limiting example 400 in which an example card frame presenting elements of a particular type is shown as determined based on element type preferences, according to some embodiments of the present disclosure. In example 400, card frame 410 through card frame 430 all display the picture content items. Where card frame 410 presents a "Title" (Title) type of content element, card frame 420 presents a "Title + brief" (Abstract) type of content element, and card frame 430 presents a "Title + brief + hot comment" type of content element. In this example, it is assumed that the user's element type preferences at the same time will not change as the content source changes,

It should be appreciated that in such examples, the final particular element presentation on the card may also be affected by the layout of the card as will be described later. For example, in the final filling of card frame 410 and card frame 420, the corresponding popular reviews may be additionally filled in

portions

415 and 425 that fill the primary recommended content item, while other portions do not display the popular reviews.

In some embodiments, based on the element of the profile type that should be presented, computing device 130 may extract the sentence from the body that is most attractive to the user as the profile element based on the content preference. In such embodiments, the computing device 130 may use the tags indicating the user content preferences to match at least one keyword from the text and use the knowledge-graph to match the associated keywords. In such some embodiments, computing device 130 may also use syntactic industry keyword recognition rules to identify keywords based on content preferences, e.g., keywords related to revenue for financial fund class preferences, keywords such as score/score for sports class preferences. Based on the plurality of keywords that are matched or identified, the computing device 130 may extract sentences from the text of the content item that are most likely to be of interest to the user as profile elements.

Extracting profiles in this way, even if the same piece of content item is recommended for application cards of different users, the content elements eventually displayed on the cards may differ due to differences in user preferences. For example, based on content preferences prompting a user for ball star A, sentences of goals and/or scores related to ball star A may be extracted from the content item text as profiles for the user. Based on another user's interest in ball star B, sentences of goal and/or ratio classification associated with ball star B may be extracted as profiles for that other user.

Similarly, in some embodiments, for an element of a picture type, computing device 130 may select a picture including an object of interest (e.g., a person) of the user as a filler element based on an indication of content preference. In some embodiments, computing device 130 may also or instead crop the picture to present the object of interest centrally.

In some embodiments, for a cover picture element of a video content item, the computing device 130 may extract frames including objects of interest to the user as covers based on an indication of content preference. For example, computing device 130 may extract key frames in which people of interest to the user are located by face recognition and combining layout evaluation with aesthetic evaluation. For example, when a user is interested in Mars A, computing device 130 may select key frames that include a shot of Mars A, and aesthetic score those frames in conjunction with a visual thermodynamic diagram and select the cover picture that is the highest scoring as the push video.

Fig. 5 shows a schematic diagram of a non-limiting example 500 in which, for different users, the computing device 130 determines different elements of the same content item for presentation on an application card, according to some embodiments of the present disclosure. In example 500, a first user is interested in ball star a, a second user is interested in ball star B, and ball star a and ball star B are included in an original picture 510 of the content item. For an application card presented to the first user, computing device 130 clips the original picture centered on sphere a. For the application card presented to the second user, computing device 130 clips the original picture centered on ball star B, as shown by XX and XX, respectively. Similarly, using the aforementioned profile extraction method, for example, computing device 130 extracts sentences associated with Mars A and Mars B as profiles for the two users, respectively.

As a result, the presentation of the content item is shown at 540 on the application card that is ultimately presented to the first user. The presentation of the content item is shown at 550 on the application card that is ultimately presented to the second user. These two presentation results, although both recommendations for the same content item, include different detail drawings and different profiles that focus on ball star a and ball star B, respectively.

In this way, content elements presented on respective cards may differ even if the same content item is pushed to different users. Therefore, the same news can be provided with the emphasis display points which meet the preference of the corresponding user for different users, so that the entrance effect of the card is enhanced.

In some embodiments, the computing device 130 may determine at least one of the aforementioned layout style and visual style in the application card as the content element to be populated based on the formal preference. The computing device 130 may then type the content elements according to the determined layout style and/or visual style and cause the type-set content elements to be presented on a card of a corresponding application (e.g., application card 140) on the client device.

FIG. 6 illustrates a schematic diagram of a non-limiting example 600 in which example card layout styles determined based on layout style preferences are shown, according to some embodiments of the present disclosure. In this style, layout style 610 is a tabular layout in which each column presents one content item, and the type of graphic is left-hand and right-hand. Layout style 620 is also a list layout in which two content items are presented per column, and the type of graphic is a context of the graph.

Further, the layout style 630 is a "content card+list" layout, with the content card portion 631 located above and the list portion 632 located below. In the content card part, the filling text is embedded into the picture, and in the list part, the picture-text type of the content item is left picture and right text. Layout style 640 is also a "content card + list" layout. Unlike layout style 630, the filler text in the location of the content card is placed outside the picture.

In some embodiments, the computing device 130 may obtain templates or typesetting rules for the respective layout styles and/or visual styles based on the layout style preferences and/or visual style preferences included in the form preferences. For example, the computing device 130 may randomly allocate each content item at each layout position and populate with corresponding elements of the content item. For another example, the computing device 130 may populate presentation elements of the content item according to preset rules. In such examples, the computing device 130 may rank the content items by the number of user content preference tags that each content item contains. The computing device 130 may then sequentially populate the content items into the style according to the fill order of the typeset positions defined by the layout style.

It should be understood that all or part of the actions in the above embodiments may be used in combination in various suitable ways. As an example, fig. 7 shows a schematic diagram 700 of data information and its relationships involved in an example process of presenting recommended content according to some embodiments of the present disclosure. This example process may be performed, for example, by the computing device 130 as shown in fig. 1, to cause recommended content to be presented on the card 140. Fig. 7 is described below in connection with the example environment 100 of fig. 1. The process is shown by way of example only, and wherein the acts are not necessarily performed in the exact order described.

Upon determining the current form preference 705 and the current content preference 710, the computing device 130 may continue to perform the example process associated with FIG. 7. As shown, the computing device may determine the visual style 720, typesetting style 725, element type 730, and media type 735 of the content that should be presented on the card 140 at this time based on the current form preferences. Based on the media type 735 and the current content preferences 710, the computing device 130 may determine content items 745 to recommend that are consistent with both from the content pool 740. Although not shown, the number of content items 745 may depend on the number of content items that can be supported by the typesetting style 725.

Further, based on the element type 730, the current content preference 710, and the typesetting style 725, the computing device 130 may extract the corresponding content element from the content item 745. The computing device 130 then performs intelligent composition based on the visual style 720 and the composition style 725 to populate the content elements 750 on the card. After determining the ordered card content 755, the computing device may cause the card content 755 to be loaded onto the card 140, resulting in a loaded card 760 that is visible to the user 110.

Using methods according to embodiments of the present disclosure to determine and present recommended content for application cards, in one aspect, when presenting recommended cards for the same application to different users in the same environment in embodiments where immediate preferences are determined based on data for a particular user, at least a portion of the recommended cards may differ in form and/or content from user to user preferences. On the other hand, for the same user, as the preferences thereof vary with the current environment, under different environmental conditions, at least a portion of the content theme and form of the card presented to the user also varies with the instant preferences, so that the presented content always conforms more to the instant preferences, thereby increasing the likelihood that the recommended content will be of interest to the user.

By way of illustration, fig. 8 shows an interface schematic of a non-limiting example 800 according to some embodiments of the present disclosure. The interface schematic shows recommended content presented in different scenarios for a card of the same application on the user device of the same user, the card being added on the negative screen of the user device. In this simplified example, the content preferences of the user are set to be substantially the same in different scenarios. For ease of illustration, this example will be described below in the context of computing device 130 of fig. 1 performing actions.

As shown in interface 810, at 8 am, the user refreshes the negative one-screen application card on the way to work. Based on various environmental information gathered by the user device (e.g., the scene is "in-car" and the device has speakers connected) and the preference model determined from the user data, the computing device 130 determines the user's immediate preferences as follows: the media type is audio, the element type is title, the layout style is a plain text list, and several topics such as "S team" are of interest.

According to the style templates of the plain text list, 5 pieces of news should be presented, the computing device 130 may obtain 5 pieces of news associated with the topic of interest from the content pool of the application and extract its title. The computing device 130 may then type the title according to the style template and cause the type-set content to be presented on the application card, as shown by card 815.

As shown in interface 820, 9 am is 28 minutes, and the user views the negative one screen during company rest time. Based on the context information and the user preference model, such as the context "in company," the computing device 130 determines the user's immediate preferences as follows: the media type is a picture, the element type is a "title+brief introduction", the layout style is multi-picture and text (the style template of which supports presentation of 3 content items), and is of interest to several topics such as "S-team".

The computing device 130 may then determine 3 news items based on the content preferences, extract titles, and extract profiles and picture cuts from the text that may be of interest to the user based on the user's current content preference tags. The computing device 130 may then typeset the extracted content items according to the corresponding style templates and cause the typeset recommended content to be presented on the application card, as shown by card 825.

It should be appreciated that the above examples are by way of illustration only and not limitation, e.g., in the two example interfaces of FIG. 8, since the interval is short and the user's content preferences do not change, the content items presented on the card are the same, but as new news enters the content pool or the user's content preferences change in different scenarios, the computing device 130 may recommend different news as well.

The user's preferences may be determined based on historical data of the user's previous browsing of application content. Typically, the data of the application card originates from an associated application (e.g., news information application). User behavior data reflecting behavior characteristics of a user operation may be collected, for example, via a user device, under the precondition of user authorization when the user performs an operation (e.g., views or clicks on a piece of content), as well as device environment data reflecting the environment in which the device is located when the operation occurs. In addition, profile data generated when a user registers or updates personal settings may also be collected.

This information collected for a particular user may be used to determine a preference model for the user. Because the user behavior records of different users are different, the obtained preferences of different users are larger, and different results are output when the instant preferences are predicted, so that different interfaces with differences of different users are further achieved. Table 6 shows examples of types of user data that may be gathered and descriptions thereof. It should be understood that the information in table 1 is shown by way of illustration and is not exhaustive.

Table 6: user data example

In some embodiments, computing device 130 may obtain a set of records about user 120, each record in the set of records including behavioral data and environmental data. The behavior data describes characteristics of a respective behavior of the user 120 accessing the content (e.g., behavior characteristics of the user 120 when accessing an application associated with the card 140), and the environment data describes characteristics of an environment in which a user device (e.g., user device 110) for performing the respective behavior is located when the respective behavior is performed. On this basis, computing device 130 may extract historical correlations between behavior data and environmental data in the set of records. The historical correlations may be used as correlations to predict the instant preferences of the user 120.

For example, the computing device 130 may convert each record into a sample pair of a set of preference tags and a set of environment tags. These sample pairs can be used as data samples for building preference models. The preference model may then be trained based on a certain sample size. The model so trained may reflect particular preferences of the user 120 when previously browsing content in a particular environment. Specific training of the model may be performed according to a general machine learning method or the like, to which embodiments of the present disclosure are not limited.

In some embodiments, the computing device 130 also considers user profile data as described above when determining the user's immediate preferences. The consideration may be incorporated into the user preference model in a suitable form. For example, the computing device 130 may convert the user's current profile data in the corresponding record into a portion of the preference label during model training.

FIG. 9 illustrates a schematic diagram of a training and use example 900 of a preference model according to some embodiments of the disclosure. The example 900 may be performed, for example, by the computing device 130 based on data for the user 120, the example 900 being described below in connection with the environment 100 of fig. 1.

Block 910 illustrates an example training process for a preference model. In this example, the computing device 130 may convert 914 the device environment data 911 in the collected records of the user 120 into an environment tag 916 and convert 912 the user behavior data 912 and the user profile data 915 into a preference tag 917. It should be appreciated that each user record may be converted into a type of label such as the media type label, the content type label, the layout label, and the visual style label described above, and will not be described again. Each tag type can be subdivided into more fields. The number of tags for the same type of formal preference may be different. Computing device 130 then trains 918 preference model 919 using sample pairs of respective environmental tags 916 and preference tags 917 for predicting instant interest in the application card (e.g., card 140) of user 120 when recommending content thereto.

Block 920 illustrates an example process for predicting using a trained preference model. In response to receiving the indication to refresh the application card 140, the computing device 130 may convert the relevant data of the environment in which the current user device 110 is located into a current environment label 921 and provide the current environment label 921 to the trained preference model 919 as input. The preference model 919 then performs predictions 919 and outputs current form preferences 922 and current content 923 of the user 120. The determined current form preferences 922 and current content 923 are then further used to determine and present recommended content for the application card as previously described and not described in further detail herein.

Over time, some of the preferences of the same user may change gradually. For example, a user may prefer football for a long period of time, but pay attention to different events for a short period of time. For example, the user may gradually transition from favorites sports to favorites football. In such embodiments, after determining the preference model using the first set of records that is the first set of records and that was obtained during the first period of time, the computing device 130 may also obtain a second set of records for the user during a second period of time that is subsequent to the first period of time. The computing device 130 may then update the determined correlations based on the historical correlations between the behavioral data and the environmental data in the second set of records.

In other words, the computing device 130 may update the preference model based on the newly collected user data in an appropriate manner such that the preference model may reflect the correlation between the updated user preferences and the environment. For example, the computing device 130 may use the new data to fine tune the existing model, or scroll out the oldest training data and add new training data, and so on. Embodiments of the present disclosure are not limited by the particular manner of updating the model. Therefore, when the content recommendation is performed on the application card of the user, the preference evolution of the user can be considered, so that the recommended content is more targeted.

Fig. 10 illustrates a schematic block diagram of an apparatus 1000 for presenting recommended content according to some embodiments of the disclosure. Apparatus 1000 may be implemented as or included in computing device 130 of fig. 1. In some embodiments, the apparatus 1000 may also be included in the user equipment 110 of fig. 1. The apparatus 100 may include a plurality of modules for performing corresponding actions in, for example, the method 200 discussed in fig. 2.

As shown in fig. 10, the apparatus 1000 includes: a receiving module 1010 configured to receive an indication of presenting first recommended content of an application on an application card of a user device; a determining module 1020 configured to determine, in response to receiving the indication, a first preference comprising a content preference and a formal preference for presentation of the content, the first preference being associated with a current environment in which the user device is located, and the first preference being different from a second preference for presentation of second recommended content within an application interface of the application upon receiving the indication; and a presentation module 1030 configured to cause presentation of first recommended content on the application card in a manner consistent with the first preference, wherein the first recommended content is determined based on the first preference, the first recommended content comprising content elements of a set of content items that are presented on the application card as representations of the set of content items.

In some embodiments, the apparatus 1000 further comprises: and a second presentation module configured to cause presentation of second recommended content within the recommended page in a manner consistent with the second preference if the recommended page within the application interface is detected to be opened upon receipt of the indication.

In some embodiments, the presentation module comprises: a media type determination module configured to determine a media type of the first recommended content based on the formal preference; and a content item determination module configured to determine a set of content items associated with the content preference and having a media type as a set of content items.

In some embodiments, the presentation module further comprises: an element type determination module configured to determine an element type of the content element based on the formal preference; and a content element determination module configured to extract, based on the content preferences and the element types, content elements associated with the content preferences and having the element types from the content items in the determined set of content items.

In some embodiments, the apparatus 1000 further comprises: a composition determination module configured to determine at least one of a layout style and a visual style of the content elements in the application card based on the form preference; a typesetting module configured to typeset the content elements according to at least one of the determined layout style and visual style; and the presentation module includes a card loading module configured to cause the typeset content elements to be presented on the card.

In some embodiments, the determining module comprises: a user determination module configured to determine a user with whom the application is currently associated on the user device; and a preference determination module configured to determine a first preference for a feature of the current environment based on a correlation between a historical behavioral characteristic of the user accessing the content and a historical environmental characteristic.

In some embodiments, the preference determination module includes: the comprehensive determination module determines preferences for the current environment based on the correlation and profile information of the user.

In some embodiments, the apparatus 1000 further comprises: an acquisition module configured to acquire a set of records about a user, wherein the records in the set of records include behavior data describing characteristics of respective behaviors of the user accessing the content and environment data describing characteristics of an environment in which a user device for performing the respective behaviors is located when the respective behaviors are performed; and a relationship determination module configured to determine a correlation by extracting historical correlations between the behavioral data and the environmental data in the set of records.

In some embodiments, the set of records is a first set of records and is acquired during a first period of time, and the apparatus 1000 further comprises: a second acquisition module configured to acquire a second set of records of the user for a second period of time after the first period of time; and a relationship updating module configured to update the determined correlation based on the historical correlation between the behavior data and the environmental data in the second set of records.

FIG. 11 shows a schematic block diagram of an example device 900 that may be used to implement embodiments of the present disclosure. Device 900 may be used to implement the functionality of computing device 130 or user device 110 shown in fig. 1. As shown, the device 1100 includes a computing unit 1101 that can perform various suitable actions and processes in accordance with computer program instructions stored in Random Access Memory (RAM) 1103 and/or Read Only Memory (ROM) 1102 or loaded from storage unit 1108 into RAM 1103 and/or ROM 1102. In the RAM 1103 and/or ROM 1102, various programs and data required for the operation of the device 1100 can also be stored. The computing unit 1101 and the RAM 1103 and/or the ROM 1102 are connected to each other by a bus 1104. An input/output (I/O) interface 1105 is also connected to bus 1104.

Various components in device 1100 are connected to I/O interface 1105, including: an input unit 1106 such as a keyboard, a mouse, etc.; an output unit 1107 such as various types of displays, speakers, and the like; a storage unit 1108, such as a magnetic disk, optical disk, etc.; and a communication unit 1109 such as a network card, modem, wireless communication transceiver, or the like. The communication unit 1109 allows the device 1100 to exchange information/data with other devices through a computer network such as the internet and/or various telecommunication networks.

The computing unit 1101 may be a variety of general purpose and/or special purpose processing components having processing and computing capabilities. Some examples of computing unit 1101 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 1101 performs the various methods and processes described above, such as method 200. For example, in some embodiments, the method 200 may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as the storage unit 1108. In some embodiments, some or all of the computer programs may be loaded and/or installed onto device 1100 via RAM and/or ROM and/or communications unit 1109. One or more of the steps of the method 200 described above may be performed when a computer program is loaded into RAM and/or ROM and executed by the computing unit 1101. Alternatively, in other embodiments, the computing unit 1101 may be configured to perform the method 200 by any other suitable means (e.g., by means of firmware).

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof, and when implemented in software, it may be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions which, when loaded and executed on a server or terminal, fully or partially produce a process or function in accordance with embodiments of the present application. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by a wired (e.g., coaxial cable, fiber optic, digital subscriber line), or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer readable storage medium may be any available medium that can be accessed by a server or terminal or a data storage device such as a server, data center, etc. that contains an integration of one or more available media. The usable medium may be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape, etc.), an optical medium (such as a digital video disk (digital video disk, DVD), etc.), or a semiconductor medium (such as a solid state disk, etc.).

Moreover, although operations are depicted in a particular order, this should be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous. Likewise, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the present disclosure. Certain features that are described in the context of separate embodiments can also be implemented in combination in a single implementation. Conversely, various features that are described in the context of a single implementation can also be implemented in multiple implementations separately or in any suitable subcombination.

Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are example forms of implementing the claims.

Claims

1. A method of presenting recommended content, comprising:

Receiving an indication of presenting first recommended content of an application on an application card of a user device;

in response to receiving the indication, determining a first preference comprising a content preference and a formal preference for content presentation, the first preference being associated with a current environment in which the user device is located and the first preference being different from a second preference for presenting second recommended content within an application interface of the application when the indication is received; and

causing the first recommended content to be presented on the application card in a manner consistent with a first preference, wherein the first recommended content is determined based on the first preference, the first recommended content comprising content elements of a set of content items that are presented on the application card as representations of the set of content items.

2. The method as recited in claim 1, further comprising:

if the recommended page in the application interface is detected to be opened when the indication is received, the second recommended content is presented in the recommended page in a manner consistent with the second preference.

3. The method of claim 1, wherein the formal preferences include preferences for one or more of:

Recommending the media type of the content;

an element type representing the content element being presented as recommended content;

a layout style in which recommended content is presented on an application card; and

the recommended content is presented in a visual style on the application card.

4. The method of claim 1, wherein determining the first recommended content based on the first preference comprises:

determining a media type of the first recommended content based on the formal preference; and

a set of content items associated with the content preference and having the media type is determined as the set of content items.

5. The method of claim 4, wherein determining the first recommended content based on the first preference further comprises:

determining an element type of the content element based on the formal preference; and

based on the content preference and the element type, content elements associated with the content preference and having the element type are extracted from the determined content items in the set of content items.

6. The method as recited in claim 1, further comprising:

determining at least one of a layout style and a visual style of the content element in the application card based on the form preference; and

Typesetting the content element according to the determined at least one of the layout style and the visual style; and is also provided with

Causing the first recommended content to be presented on the application card in a manner consistent with the first preference includes:

causing the typeset content elements to be presented on the application card.

7. The method of claim 1, wherein determining the first preference comprises:

determining a user with whom the application is currently associated on the user device; and

the first preference is determined for a feature of the current environment based on a correlation between historical behavioral characteristics and historical environmental characteristics of the user access content.

8. The method of claim 7, wherein determining the first preference for the current environment comprises:

the preference is determined for the current environment based on the correlation and profile information of the user.

9. The method of claim 7, further comprising determining the correlation by:

obtaining a set of records about the user, wherein records in the set of records include behavior data describing characteristics of respective behaviors of the user accessing content and environment data describing characteristics of an environment in which a user device for executing the respective behaviors is located when the respective behaviors are executed; and

And determining the correlation by extracting historical correlation between the behavior data and the environment data in the group of records.

10. The method of claim 9, wherein the set of records is a first set of records and is acquired during a first period of time, and further comprising:

acquiring a second set of records for the user at a second time period subsequent to the first time period; and

updating the determined correlation based on historical correlations between behavior data and environmental data in the second set of records.

11. An apparatus for presenting recommended content, comprising:

a receiving module configured to receive an indication of first recommended content of an application presented on an application card of a user device;

a determining module configured to determine, in response to receiving the indication, a first preference comprising a content preference and a formal preference for content presentation, the first preference being associated with a current environment in which the user device is located, and the first preference being different from a second preference for presenting second recommended content within an application interface of the application when the indication is received; and

A presentation module configured to cause the first recommended content to be presented on the application card in a manner consistent with a first preference.

12. An electronic device comprising a processor and a memory having stored thereon computer instructions that, when executed by the processor, cause the electronic device to perform the method of any of claims 1 to 10.

13. A computer readable storage medium having stored thereon a computer program which when executed by a processor implements the method according to any of claims 1 to 10.

14. A computer program product comprising computer executable instructions which when executed by a processor implement the method of any one of claims 1 to 10.