CN115080835A

CN115080835A - Information recommendation method and device, user side and equipment

Info

Publication number: CN115080835A
Application number: CN202110261784.3A
Authority: CN
Inventors: 蒋孜文; 龚禹; 刘青文; 欧文武; 冯宇飞; 徐国欢; 黄绍澎; 陈乐�
Original assignee: Alibaba Singapore Holdings Pte Ltd
Current assignee: Alibaba Singapore Holdings Pte Ltd
Priority date: 2021-03-10
Filing date: 2021-03-10
Publication date: 2022-09-20

Abstract

The embodiment of the application provides an information recommendation method, an information recommendation device, a user side and equipment. The method comprises the steps that a user side sends a first request to a server side, so that the server side determines at least one piece of recommendation information based on first user behavior data; the user side acquires and stores the at least one piece of recommendation information provided by the server side; and the user side updates the current recommendation result in the recommendation page at least according to the stored recommendation information. The technical scheme provided by the embodiment of the application ensures the accuracy of the recommendation result and improves the utilization rate of the computing resources of the user side.

Description

Information recommendation method and device, user side and equipment

Technical Field

The embodiment of the application relates to the technical field of networks, in particular to an information recommendation method, an information recommendation device, a user side and equipment.

Background

At present, many data processing systems providing network services, such as an object publishing platform, can provide an object-related information recommendation function, and a server can determine a plurality of pieces of recommendation information based on user behavior data in a recent period of time, generate a recommendation result, and display the recommendation result at a user side.

Due to the limitation of the user interface size, the recommendation information is usually displayed in a paging manner in the prior art. When entering a recommendation page of a user side for the first time, the user side initiates a paging request to a server side, the server side determines a recommendation result containing a predetermined number of recommendation information by combining user behavior data of a last period of time, and displays the recommendation result in the recommendation page of the client side, and a user can browse the recommendation information in the recommendation result and perform corresponding operation on the recommendation information and the like; after the recommendation information in the recommendation page is completely exposed, the user side can initiate a paging request to the server side again, the server side can continue to combine with the user behavior data to generate recommendation results containing a predetermined number of recommendation information, and the current recommendation results are displayed after the previous recommendation results are arranged in the recommendation page of the user side.

As can be seen from the above description, the user side is only responsible for sending the paging request and displaying the recommendation result, and the computing resources of the user side are not fully utilized, which is a waste of the computing resources of the user side.

Disclosure of Invention

The embodiment of the application provides an information recommendation method, an information recommendation device, a client and equipment, and aims to solve the technical problem that in the prior art, computing resources of the user side are not fully utilized.

In a first aspect, an embodiment of the present application provides an information recommendation method, which is applied to a user side, and the method includes:

sending a first request to a server, so that the server determines at least one piece of recommendation information based on first user behavior data;

acquiring the at least one piece of recommendation information provided by the server and storing the at least one piece of recommendation information;

and updating the current recommendation result in the recommendation page at least according to the stored recommendation information.

In a second aspect, an embodiment of the present application provides an information recommendation method, which is applied to a server, and the method includes:

receiving a first request sent by a user side;

determining at least one recommendation based on the first user behavior data;

and providing the at least one piece of recommendation information to the user side so that the user side can store the at least one piece of recommendation information, and updating the current recommendation result in the recommendation page at least according to the stored recommendation information.

In a third aspect, an embodiment of the present application provides an information recommendation apparatus, including:

the first request module is used for sending a first request to the server so that the server can determine at least one piece of recommendation information based on the first user behavior data;

the first acquisition module is used for acquiring the at least one piece of recommendation information provided by the server and storing the at least one piece of recommendation information;

and the updating module is used for updating the current recommendation result in the recommendation page at least according to the stored recommendation information.

In a fourth aspect, an embodiment of the present application provides a user side, which provides a display interface and a storage module;

the storage module stores at least one piece of recommendation information which is provided by the server and determined based on the first request;

and the display interface displays the current recommendation result and updates the current recommendation result at least according to the recommendation information stored in the storage module.

In a fifth aspect, an embodiment of the present application provides an information recommendation apparatus, including:

the first request receiving module is used for receiving a first request sent by a user side;

a first determining module for determining at least one recommendation based on the first user behavior data;

the first providing module is used for providing the at least one piece of recommendation information to the user side so that the user side can store the at least one piece of recommendation information, and updating the current recommendation result in the recommendation page at least according to the stored recommendation information.

In a sixth aspect, an embodiment of the present application provides an electronic device, which includes a display component, a storage component, and a processing component;

the storage component stores one or more computer instructions; the one or more computer instructions are for a processing component to invoke and execute to implement the information recommendation method according to the first aspect.

In a seventh aspect, an embodiment of the present application provides a server, including a storage component and a processing component;

the storage component stores one or more computer instructions; the one or more computer instructions are for the processing component to call and execute to implement the information recommendation method according to the second aspect.

In the embodiment of the application, a user side sends a first request to a server side, and the server side determines at least one piece of recommendation information based on first user behavior data; the user side stores the at least one piece of recommendation information provided by the server side, and updates the recommendation result in the recommendation page at least according to the stored recommendation information.

These and other aspects of the present application will be more readily apparent from the following description of the embodiments.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative efforts.

FIG. 1 is a flow chart illustrating an embodiment of a method for recommending information provided by the present application;

FIG. 2 is a flow chart illustrating a further embodiment of an information recommendation method provided herein;

FIG. 3 is an interaction diagram illustrating an information recommendation process in a practical application according to an embodiment of the present application;

FIG. 4 is a schematic structural diagram illustrating an embodiment of an information recommendation device provided in the present application;

FIG. 5 is a schematic diagram illustrating an embodiment of an electronic device provided by the present application;

FIG. 6 is a schematic structural diagram of another embodiment of an information recommendation device provided by the present application;

fig. 7 is a schematic structural diagram illustrating an embodiment of a server provided in the present application.

Detailed Description

In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application.

In some of the flows described in the specification and claims of this application and in the above-described figures, a number of operations are included that occur in a particular order, but it should be clearly understood that these operations may be performed out of order or in parallel as they occur herein, the number of operations, e.g., 101, 102, etc., merely being used to distinguish between various operations, and the number itself does not represent any order of performance. Additionally, the flows may include more or fewer operations, and the operations may be performed sequentially or in parallel. It should be noted that, the descriptions of "first", "second", etc. in this document are used for distinguishing different messages, devices, modules, etc., and do not represent a sequential order, nor limit the types of "first" and "second" to be different.

The technical scheme of the embodiment of the application can be applied to any network application scene capable of recommending information to a user, the recommendation information can refer to the related information of the recommended object, and is used for realizing recommendation and the like of the recommended object based on the recommendation information.

According to the description in the background art, the recommendation results generated based on a plurality of pieces of recommendation information are displayed in a recommendation page at present and are limited by the limitations of user side interface size and the like, specifically, the recommendation results are displayed in a paging mode, after the recommendation information entering the recommendation page for the first time and in the current recommendation interface is completely exposed, a paging request is triggered, and the server side generates the recommendation results including a predetermined number of pieces of recommendation information by combining with user behavior data in the latest period of time. In the information recommendation mode, the user side is only responsible for sending the paging request and displaying the recommendation result, and the computing resources of the client side are not fully utilized, which is a waste for the computing resources of the client side.

In order to improve the utilization rate of computing resources of a client, an inventor provides a technical scheme of the application through a series of researches, in the embodiment of the application, a client sends a first request to a server, and the server determines at least one piece of recommendation information based on first user behavior data; the user side obtains the at least one piece of recommendation information provided by the server side, stores the at least one piece of recommendation information, and can update the current recommendation result in the recommendation page at least according to the stored recommendation information. By storing the recommendation information at the user side, the user side can update the recommendation result by combining the stored recommendation information, so that the accuracy of the recommendation result can be improved.

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

Fig. 1 is a flowchart of an embodiment of an information recommendation method provided in an embodiment of the present application, where the technical solution of the embodiment is executed by a user side, and the method may include the following steps:

101: and sending a first request to the server side, so that the server side determines at least one piece of recommendation information based on the first user behavior data.

The first user behavior data includes user behavior data of a last period of time, such as user behavior data generated within a first time range, such as within 3 days, including a current time and before the current time.

The user side may send the first request to the server side based on a trigger event, and several implementation manners of the trigger event are described in detail in the following embodiments.

In addition, the service end may provide recommendation information of different information types, taking an e-commerce scenario as an example, the information types may include short videos, advertisements, commodities, and the like, and the corresponding recommendation information may be video prompt information, advertisement prompt information, commodity prompt information, and the like. Thus, optionally, sending the first request to the server may include: determining at least one information type; and sending a first request aiming at least one information type to the server side, so that the server side determines at least one piece of recommendation information corresponding to the at least one information type based on the user behavior data.

The method for determining the recommendation information by the server based on the first user behavior data may be the same as the existing method, for example, the user preference feature may be identified based on the first user behavior data, where the user preference feature may refer to an information type preferred by the user and an object feature of a recommendation object corresponding to the recommendation information, and for example, when the recommendation object is a commodity, the object feature may include a commodity type. The identification of the user preference characteristics can be realized based on a pre-trained characteristic identification model, and can also be realized in a statistical mode; thereafter, at least one piece of recommendation information may be determined based on how similar the recommendation information is to the user preference feature. The server side can also determine the arrangement sequence of the at least one piece of recommendation information according to the similarity degree with the preference characteristics of the user and the sequence from large to small of the similarity degree, and provides the at least one piece of recommendation information and the arrangement sequence thereof for the user side.

Certainly, in practical application, in order to improve recommendation accuracy, a process of determining recommendation information by the server based on the first user behavior data may be very complex, for example, stages such as recall, rough ranking, fine ranking and the like may be performed for each information type, different stages may use feature recognition models with different accuracies to recognize user preference features or use calculation models with different accuracies to calculate similarity degrees and the like of the recommendation information and the user preference features, then, based on the similarity degrees, screening and sorting of the recommendation information are performed, and based on fine ranking results of different information types, mixed ranking and the like may also be performed.

102: and acquiring the at least one piece of recommendation information provided by the server and storing the at least one piece of recommendation information.

The at least one piece of recommendation information may be stored in a cache, the cache may be a persistent cache, and the cache capacity may be preset in combination with an actual demand and to ensure storage stability of the user side.

103: and updating the current recommendation result in the recommendation page at least according to the stored recommendation information.

Wherein the current recommendation result may be generated by the server based on the latest paging request; in addition, under the condition that the first request may be sent multiple times, the current recommendation result may also be obtained after the user side updates the recommendation result shown in the recommendation page at least the last time according to the stored recommendation information.

Optionally, unexposed recommendation information in the current recommendation result in the recommendation page may be updated at least according to the stored recommendation information. The exposure of the recommendation information may mean that the recommendation information is displayed on a user-side display interface for a user to browse and perform further operations.

In this embodiment, the user side obtains the at least one piece of recommendation information determined by the server side based on the first user behavior data by sending the first request to the server side, and stores the at least one piece of recommendation information, so that the current recommendation result can be updated in combination with the stored recommendation information, the accuracy of the recommendation result can be improved, the computing resources of the user side are utilized, and the utilization rate of the computing resources of the user side is improved.

The server may specifically determine a first number of pieces of recommendation information based on the first user behavior data, and in some embodiments, acquiring at least one piece of recommendation information provided by the server and storing the at least one piece of recommendation information may include:

acquiring a first quantity of recommendation information provided by a server;

based on the second user behavior data, sorting the first number of pieces of recommendation information and the stored recommendation information to obtain a first sorting result;

and selecting a second amount of recommendation information to store according to the first sequencing result.

Optionally, a persistent cache may be set in the user side, and the recommendation information obtained from the server side may be stored in the cache, so that based on the second user behavior data, the first number of recommendation information and the recommendation information in the cache may be sorted to obtain a first sorting result; then, from the first sorting result, a second number of pieces of recommendation information are selected to update the recommendation information stored in the cache, and the like.

The first sorting result includes the sorting order of the first number of pieces of recommendation information and the stored recommendation information, and the like.

The second quantity may be determined by combining the storage capacity of the user side, and may be determined according to the cache capacity when the user side stores the recommendation information in the cache, and the second quantity may be the maximum quantity that the cache can store.

In this embodiment, the user side may combine the second user behavior data to screen and sort the first number of pieces of recommendation information and the stored recommendation information, and reselect the second number of pieces of recommendation information to store, so as to update the stored recommendation information, and ensure that the stored recommendation information is more accurate, that is, more in line with the user intention, so as to further improve the accuracy of the recommendation result.

The second user behavior data may be, for example, user behavior data generated in a second time range including the current time and a time before the current time. The first time range may be the same as or different from the second time range, and the second user behavior data is, compared with the first user behavior data, the latest user behavior data in the first user behavior data is earlier than the latest user behavior data in the second user behavior data, that is, the second user behavior data is more real-time and includes the latest user behavior data, because the user can perform corresponding operations, such as clicking, collecting, forwarding, or viewing, on the recommendation information in the recommendation result in the browsing process of the current recommendation result, so as to generate a user behavior, especially for a user with a high interaction demand, the exposure duration of the current recommendation result is generally longer, such as several minutes, compared with the current time, the recommendation result in the recommendation page is determined by combining the previous user behavior data for several minutes, the latest user behavior data generated in the last few minutes is lost, and thus the recommendation result may be inaccurate. In this embodiment, the second user behavior data is combined, the recommended information is screened and sorted again, and the recommended information is reselected for storage based on the sorting result, so that the recommendation accuracy of the stored recommended information can be ensured, and the accuracy of the recommendation result can be improved.

Wherein the user preference characteristic may be first identified based on the second user behavior data, and the first ranking result may be determined according to a degree of similarity of the recommendation information and the user preference characteristic. The user side not only stores the recommendation information, but also correspondingly stores the arrangement sequence of each recommendation information and the like.

The identification mode of the user preference features and the calculation mode of the similarity degree can be realized by adopting the existing mode, and the method is not particularly limited in the present application.

After the current recommendation result is updated at least according to the stored recommendation information, if the recommendation information in the current recommendation result is replaced by the recommendation information, the number of the stored recommendation information is correspondingly reduced.

In some embodiments, updating the recommendation result in the recommendation page based at least on the saved recommendation information may include:

selecting a third number of recommendation information from the stored recommendation information according to the first sequencing result;

and updating the current recommendation result in the recommendation page based on the third amount of recommendation information.

Alternatively, the third amount may be an unexposed amount of the recommendation information in the current recommendation result. Accordingly, the method may further comprise: determining the unexposed quantity of the recommendation information in the current recommendation result as a third quantity; updating the recommendation result in the recommendation page based on the third number of recommendation information may be updating the unexposed recommendation information in the recommendation page based on the third number of recommendation information.

Wherein, the third number of recommendation information may replace the unexposed recommendation information. In addition, in practical applications, the recommendation result may be obtained by aggregating based on the recommendation information to obtain a plurality of sets of recommendation information, where the recommendation result includes the aggregation prompt information of each set of recommendation information, and the aggregation prompt information may be indexed to a page showing the recommendation information in the set, and the recommendation result may further include a display position of each aggregation prompt information in the recommendation page. The present application is not limited to a specific form of the recommendation result generated based on the plurality of pieces of recommendation information.

After the current recommendation result in the recommendation page is updated based on the third amount of recommendation information, the third amount of recommendation information may be deleted from the stored recommendation information.

sorting the stored recommendation information and the recommendation information in the recommendation page based on the third user behavior data to obtain a second sorting result;

and selecting a fourth amount of recommendation information according to the second sorting result to update the current recommendation result in the recommendation page.

As can be seen from the foregoing description, in the process of browsing the current recommendation result by the user, corresponding operations, such as clicking, collecting, forwarding, or viewing, may be performed on the recommendation information in the recommendation result, so as to generate a user behavior, especially for a user with a high interaction demand, the exposure duration of the current recommendation result is generally longer, for example, several minutes, and compared with the current time, the recommendation result in the recommendation page is determined by combining the user behavior data before several minutes, and the latest user behavior data generated in the last several minutes is lost, so the recommendation result may be inaccurate. In this embodiment, the screening and sorting of the recommendation information are performed again based on the third user behavior data. The third user behavior data may be, for example, user behavior data generated in a third time range including the current time and prior to the current time. The first time range may be the same as or different from the third time range, and the third user behavior data is, compared to the second user behavior data, earlier in occurrence time of the latest user behavior data in the first user behavior data than in the third user behavior data, that is, the third user behavior data is more real-time and includes the latest user behavior data. Therefore, the third user behavior data is utilized to re-screen recommendation information from the recommendation information in the recommendation page and the stored recommendation information to update the current recommendation result, and the accuracy of the recommendation result can be further ensured.

The fourth number may refer to an unexposed number of the recommendation information in the recommendation page, and therefore, selecting the fourth number of recommendation information to update the current recommendation result in the recommendation page may be selecting the fourth number of recommendation information to update the unexposed recommendation information in the recommendation page.

Optionally, the current recommendation result of the recommendation page may be updated by combining the stored recommendation information and the unexposed recommendation information in the recommendation page.

Therefore, based on the third user behavior data, the stored recommendation information and the recommendation information in the recommendation page are sorted, and the second sorting result may be obtained specifically as follows: and sequencing the stored recommendation information and the unexposed recommendation information in the recommendation page based on the third user behavior data to obtain a second sequencing result so as to avoid repeated recommendation information in the recommendation result and improve user experience.

As can be seen from the foregoing description, at least one information type may be determined first, and a first request for the at least one information type may be sent to the server. Therefore, the server side can specifically determine at least one piece of recommendation information corresponding to at least one information type based on the first user behavior data. Wherein each information type may correspond to one or more pieces of recommendation information.

Further, in some embodiments, the method may further comprise:

determining the number of requests corresponding to at least one information type;

sending the first request for at least one information type to the server may include: and sending a first request to the server based on the at least one information type and the request quantity respectively corresponding to the at least one information type, so that the server determines recommendation information with respective request quantity respectively corresponding to the at least one information type based on the user behavior data.

The number of requests corresponding to each information type may be preset in combination with an actual requirement, and may also be determined in combination with the number of pieces of recommendation information corresponding to the information type in the stored recommendation information, for example, a preset number of pieces of recommendation information may be stored for each information type, and the number of requests corresponding to each information type may be calculated by subtracting the remaining number of recommendation information corresponding to the information type in the stored recommendation information from the preset number corresponding to the information type.

At least one information type in the first request can be preset by combining with actual requirements;

furthermore, at least one information type preferred by the user may be determined based on the fourth user behavior data. As an alternative, at least one information type of the user preference may be identified based on the fourth user behavior data using a pre-trained information type identification model. The information type recognition model can be obtained by training based on the sample behavior data and the information type labels corresponding to the sample behavior data.

The fourth user behavior data may be, for example, user behavior data generated in a fourth time range including the current time and a time before the current time. The first time range may be the same as or different from the fourth time range,

as another alternative, the fourth user behavior data may be generated by detecting a corresponding user operation performed by the user on the current recommendation result. For example, the operation type and the operation object of the user operation can be included; the operation type may be, for example, clicking, collecting, browsing, etc., the operation object is the recommendation information, and at least one information type preferred by the user may be determined by counting the fourth row behavior data, for example, an information type in which the user's number of clicks and/or collection times is greater than a certain threshold, that is, the information type preferred by the user may be determined.

Of course, a corresponding control whether each piece of recommendation information is interested in may be correspondingly displayed in the recommendation page, the fourth user behavior data may be generated by executing a user operation for the type of control, and statistics of the fourth user behavior data may determine the information type of the recommendation information that the user is interested in, that is, may be used as the information type of the user preference.

In some embodiments, sending the first request to the server comprises: and responding to a trigger event, and sending a first request to the server.

The trigger event may be generated when any one or more of the following trigger event generation conditions are satisfied, and the trigger event generation conditions are described below:

trigger event generation condition one:

user behavior is detected that satisfies a first particular condition.

For example, the first specific condition may refer to any user behavior.

The method can be characterized in that when any user behavior is detected, a trigger event is generated, and a first request is sent to the server once.

For another example, the first specific condition may refer to a browsing operation in which a browsing speed for the current recommendation result is greater than a preset speed, for example.

The method may include detecting a browsing operation with a browsing speed higher than a preset speed for a current recommendation result, generating a trigger event, and sending a first request to the server once.

The browsing speed may refer to a swipe speed of a swipe operation performed for the current recommendation result. The browsing speed is higher than the preset speed, that is, the user is not interested in the current recommendation result, and the current recommendation result is not accurate, so that the first request can be sent to the server.

Triggering event generating condition two:

a paging trigger event is detected.

The first request may specifically be a paging request, where the paging request may be generated when a paging trigger event is detected, and the triggering event is sent to the server once.

The paging trigger event may refer to that the recommendation information in the current recommendation result is completely exposed.

Triggering event generating condition three:

the current time reaches the timing generation time of the predetermined period.

The trigger event may be generated at regular time according to a predetermined period, that is, the first request may be sent to the server at regular time according to the predetermined period.

The triggering event generating condition four:

detecting that the exposure quantity of recommendation information in the current recommendation result reaches a first quantity threshold value;

the trigger event may be generated when the trigger event generation condition is satisfied. That is, when the exposure quantity of the recommendation information in the current recommendation result reaches the first quantity threshold, a trigger event is generated, and a first request is sent to the server once. The first number threshold may be, for example, one half of the total number of pieces of recommendation information in the current recommendation result.

When the first quantity threshold is the total quantity of the recommendation information in the current recommendation result, the trigger event is also a paging trigger event.

In addition, the trigger event may be generated when both the first trigger event generation condition and the fourth trigger event generation condition are satisfied, or when both the third trigger event generation condition and the fourth trigger event generation condition are satisfied.

Triggering event generation condition five:

the stored recommendation information meeting the user preference characteristics meets the trigger condition.

The recommendation information conforming to the user preference characteristics may be recommendation information having a similarity degree with the user preference characteristics larger than a certain numerical value.

The user preference feature may be identified based on fifth user behavior data, which may include user behavior data generated at the current time and in a fifth time range before the current time. The first time range may be the same as or different from the fifth time range. Alternatively, user preference features or the like may be identified using a feature recognition model.

The trigger condition may be, for example, that the number of pieces of recommendation information that meet the user preference feature in the stored recommendation information is smaller than a preset number.

When the trigger event generation condition five is met, a trigger event is generated, and a first request is sent to the server.

Or when the triggering event generating condition five and at least one of the triggering event generating condition one, the triggering event generating condition three and the triggering event generating condition four are simultaneously satisfied, generating a triggering event and sending a first request to the server.

Trigger event generation condition six:

detecting that the number of stored recommendation information is less than a second number threshold.

When the trigger event generation condition six is met, a trigger event is generated, and a first request is sent to the server.

The trigger event may be generated and the first request may be sent to the server when at least one of the trigger event generation condition six, the trigger event generation condition one, the trigger event generation condition three, the trigger event generation condition four, and the trigger event generation condition five is simultaneously satisfied.

The trigger event generating condition seven:

and detecting that the quantity of the recommended information corresponding to any stored information type is less than a third quantity threshold value.

When the seventh condition for generating the trigger event is met, the trigger event is generated, and a first request for any information type is sent to the server.

Of course, the trigger event may be generated and the first request may be sent to the server if the trigger event generating condition seven is satisfied, and if at least one of the trigger event generating condition one, the trigger event generating condition three, the trigger event generating condition four, and the trigger event generating condition five is satisfied.

In addition, since the at least one information type may be determined in response to the trigger event, specifically, the first request for the at least one information type may be sent to the server.

The at least one information type may also be determined based on the trigger event, and the information types corresponding to different trigger events may be preset in combination with actual requirements.

Further, in some embodiments, the method may further comprise:

sending a paging request to the server so that the server can respond to the paging request to generate a current recommendation result;

acquiring a recommendation result provided by a server;

and displaying the recommendation result in a recommendation page.

The current recommendation result shown in the recommendation page may be generated by the service end based on the latest paging request, or obtained after the recommendation result shown in the recommendation page is updated last time according to at least the stored recommendation information.

Optionally, sending the paging request to the server may include:

determining one or more information types;

and sending the paging request aiming at the one or more information types to a server.

The information type corresponding to the paging request may be preset in combination with the actual demand. Therefore, under the condition that the paging request exists, the information type corresponding to the paging request can be preset according to actual requirements, and the paging request can be initiated only aiming at the information type corresponding to the paging request.

In some embodiments, the method may further comprise:

sending a first request for at least one information type to the server comprises:

and sending a first request to the server based on the at least one information type and the request quantity respectively corresponding to the at least one information type, so that the server determines recommendation information with respective request quantity respectively corresponding to the at least one information type based on the user behavior data.

The request data corresponding to each information type can be preset in combination with actual requirements.

and responding to the updating event, and updating the recommendation result in the recommendation page at least according to the stored recommendation information.

The update event may be generated when any one of the following conditions or a plurality of conditions for generating the update event are satisfied, and the conditions for generating the update event are described below:

update event generation condition a:

detecting a user behavior satisfying a second specific condition;

the second specific condition may be the same as the first specific condition.

Alternatively, the second specific condition may refer to, for example, any user behavior.

That is, any user behavior for the current recommendation result is detected, and an update event is generated.

For another example, the second specific condition may refer to a browsing operation in which a browsing speed for the current recommendation result is greater than a preset speed.

The updating event can be generated by detecting the browsing operation aiming at the current recommendation result, wherein the browsing speed is higher than the preset speed, and the first request is sent to the server once.

The browsing speed may refer to a swipe speed for a swipe operation performed when recommending a result. The browsing speed is higher than the preset speed, that is, the user is not interested in the current recommendation result, and the current recommendation result is not accurate, so that the first request can be sent to the server.

Update event generation condition B:

identifying that a user has an updating intention;

alternatively, the user presence update intention may be identified based on the sixth behavior data. Whether the user has an updating intention may be identified based on the sixth behavior data using an intention recognition model. The intention recognition model can be obtained by training in advance based on the sample behavior data and the label of whether the update intention exists or not corresponding to the sample behavior data. The sixth behavior data may be user behavior data that may include the current time and generated within a sixth time range prior to the current time. The first time range may be the same as or different from the sixth time range.

Update event generation condition C:

and detecting that the exposure number of the recommendation information in the current recommendation result reaches a fourth number threshold.

The fourth quantity threshold may be, for example, one half or two thirds of the total quantity of the recommendation information in the current recommendation result.

Fig. 2 is a flowchart of another embodiment of an information recommendation method provided by an embodiment of the present application, where the technical solution of the embodiment is executed by a server, and the method may include the following steps:

201: receiving a first request sent by a user terminal.

202: at least one recommendation is determined based on the first user behavior data.

The manner of determining the at least one piece of recommendation information based on the first user behavior data may be the same as the existing manner, and is not described herein again.

For example, a user preference feature may be identified based on the first user behavior data, where the user preference feature may refer to a type of information preferred by the user and an object feature of a recommended object corresponding to the recommendation information, and for example, when the recommended object is a commodity, the object feature may include a commodity type and the like. The identification of the user preference features can be realized based on a pre-trained feature identification model, and can also be realized in a statistical mode; thereafter, at least one piece of recommendation information may be determined based on how similar the recommendation information is to the user preference feature. The server side can also determine the arrangement sequence of the at least one piece of recommendation information according to the similarity degree with the preference characteristics of the user and the sequence from large to small of the similarity degree, and provides the at least one piece of recommendation information and the arrangement sequence thereof for the user side.

Certainly, in practical application, in order to improve recommendation accuracy, a process of determining recommendation information by the server based on the first user behavior data may be very complex, for example, stages of recalling, coarse ranking, fine ranking and the like may be performed for each information type, different stages may employ feature identification models with different accuracies to identify user preference features or employ calculation models with different accuracies to calculate similarity degrees and the like between the recommendation information and the user preference features, then screening and sorting of the recommendation information are performed based on the similarity degrees, and for fine ranking results of different information types, mixed ranking and the like may be performed.

203: and providing the at least one piece of recommendation information to the user side so that the user side can store the at least one piece of recommendation information, and updating the current recommendation result in the recommendation page at least according to the stored recommendation information.

The operation performed by the user side can be described in the foregoing embodiments, and will not be described herein again.

In some embodiments, the server may further receive a paging request sent by the user side, so as to generate a recommendation result in response to the paging request, and display the recommendation result in a recommendation page of the user side.

The current recommendation result displayed on the recommendation page by the user side can be generated by the server side based on the latest paging request, or the user side updates the recommendation result in the recommendation result at least for the latest time according to the stored recommendation information.

In a practical application, the technical scheme of the application can be applied to an electronic commerce scene, a data processing system consisting of a user side and a server side can provide a commodity transaction function, the server side can provide a commodity description page of each commodity to be displayed at the user side, and a user can know detailed information of the commodity and perform operations such as collection, shopping cart adding, purchase and the like based on the commodity description page. The service end can adopt a commodity description page for commodity introduction and can also adopt various promotion forms for commodity introduction, such as short video, live broadcast, advertisement, light application (namely small program) and the like.

In an e-commerce scenario, the recommendation information may refer to information related to a commodity. Since the server can provide content forms such as short videos, live broadcasts, advertisements, commodities (corresponding to commodity description pages) and the like to introduce commodities so as to stimulate users to purchase and the like, the server can generate recommendation information of a plurality of information types, namely, the information types can refer to the short videos, the live broadcasts, the contents, the advertisements, the commodities and the like. Since the short videos, the live broadcasts, the contents, the advertisements and the like also relate to the commodities, recommendation objects corresponding to the recommendation information are the commodities, and for different information types, the recommendation information can be indexed to content output pages corresponding to the information types, for example, the recommendation information of the short videos can be linked to playing pages corresponding to the short videos, the recommendation information of the live broadcasts can be linked to playing pages corresponding to the live broadcasts, the recommendation information of the advertisements can be linked to display pages corresponding to the advertisements, and the recommendation information of the commodities can be linked to corresponding commodity description pages and the like.

The method includes the steps that a plurality of pieces of recommendation information are displayed on a recommendation page, in a traditional mode, a server side determines the plurality of pieces of recommendation information based on user behavior data in a recent period of time, a recommendation result is generated based on the plurality of pieces of recommendation information, and the recommendation result is specifically displayed in the recommendation page. The recommendation page may refer to a home page of the user side. The traditional recommendation mode adopts a paging request mechanism, and a user side is only responsible for sending paging requests and displaying recommendation results in recommendation pages, which is a waste for computing resources of the user side.

And after the server receives the paging request, a recommendation result is generated based on the user behavior data in the last period of time, the paging request is triggered again only after the current recommendation result is completely exposed, if the exposure duration of the current recommendation result is longer, the user behavior data can be generated in the exposure duration, the decision of the server on the current recommendation result can lack the user behavior data which is generated recently, so that the recommendation result is inaccurate, and if the exposure duration of the current recommendation result is shorter, the paging request can be frequently sent, the generated recommendation results are not very different, which is a waste for the server computing resources. In addition, because the server may provide recommendation information of multiple information types, the importance levels of the information types are different, in a conventional manner, after the server receives a paging request, the server performs screening, sorting and the like on the recommendation information of the information type based on user behavior data for each information type, and finally performs a mixing operation such as mixing and sorting on the recommendation information of the multiple information types, the recommendation information of each information type consumes the same computing resources of the server, and the computing resources of the server are wasted.

With the technical solution of the embodiment of the present application, as shown in fig. 3, a first request mechanism may be used for recommendation, and the user side 301 may make a request decision, and in response to a trigger event, may first determine at least one information type, and send a first request for the at least one information type to the server side 302. For example, assume that the information types include merchandise, short videos, live broadcasts, advertisements, and the like. The at least one type of information may include goods and advertisements. By making the request decision to initiate the first request only for at least one information type, the waste of computing resources of the server can be reduced. Several possible generation manners of the trigger event are described in detail above, and are not described herein again.

After receiving the first request, the server 302 may determine at least one piece of recommendation information corresponding to the at least one information type based on the current user behavior data, where a response operation of the server 302 to the first request may be the same as a response operation to the page request, and details are not described here again.

The server 302 may provide at least one recommendation information to the user 301, and the user 301 may store the at least one recommendation information in the cache 30.

The user side 301 may respond to the update event, sort and filter the recommendation information in the cache and the unexposed recommendation information in the current recommendation result based on the current user behavior data again, that is, rearrange on the side, and select a certain amount of recommendation information to update the unexposed recommendation information in the current recommendation result in the recommendation page 31. The details of several possible generation manners of the update event can be found in the foregoing, and are not described herein again.

By performing request decision and on-end rearrangement on the user side, the accuracy of the recommendation result can be improved, the utilization rate of the computing resources of the user side is improved, and the waste of the computing resources is avoided.

The method and the device for generating the recommendation result can add a first request mechanism under an original paging request mechanism, and certainly can only adopt the first request mechanism to generate the recommendation result for all information types or partial information types. Under the condition that all information types or part of information types are recommended only by adopting the first request mechanism, the calculation pressure of the calculation resources of the server side can be relieved to a certain extent, and the accuracy of the recommendation result can be ensured by combining the recommendation operation of the user side.

Fig. 4 is a schematic structural diagram of an embodiment of an information recommendation device provided in the present application, where the device may include:

the first request module 401 is configured to send a first request to the server, so that the server determines at least one piece of recommendation information based on the first user behavior data;

a first obtaining module 402, configured to obtain at least one piece of recommendation information provided by a server and store the at least one piece of recommendation information;

an updating module 403, configured to update the current recommendation result in the recommendation page at least according to the stored recommendation information.

In some embodiments, the server determines a first number of pieces of recommendation information based on the user behavior data;

the first obtaining module is specifically used for obtaining a first number of pieces of recommendation information provided by the server; based on the second user behavior data, sorting the first number of pieces of recommendation information and the stored recommendation information to obtain a first sorting result; and selecting a second amount of recommendation information to store again according to the first sequencing result.

In some embodiments, the updating module is specifically configured to select a third number of pieces of recommendation information from the stored pieces of recommendation information according to the first sorting result; and updating the current recommendation result in the recommendation page based on the third number of recommendation information.

In some embodiments, the apparatus may further comprise:

the quantity calculation module is used for taking the unexposed quantity of the recommendation information in the current recommendation result as a third quantity;

the updating module is specifically configured to update unexposed recommendation information in the recommendation page based on the third number of recommendation information.

In some embodiments, the update module may be specifically configured to rank, based on the third user behavior data, the stored recommendation information and recommendation information in the recommendation page, and obtain a second ranking result; and selecting a fourth quantity of recommendation information to update the current recommendation result in the recommendation page according to the second sequencing result.

In some embodiments, the first request module may be specifically configured to determine at least one information type; and sending a first request aiming at least one information type to the server side, so that the server side determines at least one piece of recommendation information corresponding to the at least one information type based on the first user behavior data.

In some embodiments, the first request module is further configured to determine a number of requests corresponding to each of the at least one information type;

the first request module sends a first request for at least one information type to the server side, and the first request module comprises the following steps: and sending a first request to the server based on the at least one information type and the request quantity respectively corresponding to the at least one information type, so that the server determines recommendation information with respective request quantity respectively corresponding to the at least one information type based on the user behavior data.

In some embodiments, the first request module may be specifically configured to send the first request to the server in response to a trigger event.

In some embodiments, the apparatus may further comprise:

a first generation module, configured to generate a trigger event if one or more trigger event generation conditions are met, where the one or more trigger event generation conditions include:

detecting a user behavior satisfying a first specific condition;

detecting a paging trigger event;

the current time reaches the timing generation time of a preset period;

recommending information which accords with the user preference characteristics in the stored recommending information meets the triggering condition;

detecting that the number of the stored recommendation information is smaller than a second number threshold;

and detecting that the quantity of the recommended information corresponding to any stored information type is smaller than a third quantity threshold value.

In some embodiments, the update module is specifically configured to update the recommendation result in the recommendation page in response to the update event based at least on the stored recommendation information.

In some embodiments, the apparatus may further comprise:

a second generation module, configured to generate an update event if one or more update event generation conditions are met, where the one or more update event generation conditions include:

detecting a user behavior satisfying a second specific condition;

identifying that a user has an updating intention;

In some embodiments, the apparatus may further comprise:

the second request module is used for sending a paging request to the server so that the server can respond to the paging request and generate a current recommendation result;

the second acquisition module is used for acquiring the current recommendation result provided by the server;

and the display module is used for displaying the current recommendation result in the recommendation page.

Optionally, the second request module may be specifically configured to determine one or more information types; and sending a paging request aiming at one or more information types to a server.

The information recommendation apparatus shown in fig. 4 may execute the information recommendation method shown in the embodiment shown in fig. 1, and the implementation principle and the technical effect are not repeated. The detailed description of the information recommendation device in the above embodiments has been given in the embodiments related to the method, and the detailed description will not be given here.

In addition, the application provides a user side, which can provide a display interface and a storage module; optionally, the storage module may be a cache, and specifically may refer to a persistent cache provided by a user side;

and displaying the current recommendation result on the display interface, and updating the current recommendation result at least according to the recommendation information in the storage module.

In one possible design, the information recommendation apparatus in the embodiment shown in fig. 4 may be implemented as an electronic device, which may be, for example, a personal computer, a mobile phone, a tablet computer, other portable terminals, and the like in practical applications, and as shown in fig. 5, the electronic device may include a display component 501, a storage group 502, and a processing component 503;

the storage component 502 stores one or more computer instructions for the processing component 503 to call and execute to implement the information recommendation method as shown in fig. 1.

The display component 501 provides a display interface to display a recommendation page, and displays the current recommendation result in the recommendation interface.

Of course, the electronic device may of course also comprise other components, such as input/output interfaces, communication components, etc.

The input/output interface provides an interface between the processing components and peripheral interface modules, which may be output devices, input devices, etc.

The communication component is configured to facilitate wired or wireless communication between the electronic device and other devices, and the like.

An embodiment of the present application further provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a computer, the information recommendation method in the embodiment shown in fig. 1 may be implemented.

Fig. 6 is a schematic structural diagram of another embodiment of an information recommendation device provided in an embodiment of the present application, where the device may include:

a first receiving module 601, configured to receive a first request sent by a user side;

a first determining module 602, configured to determine at least one piece of recommendation information based on the first user behavior data;

the first providing module 603 is configured to provide the at least one piece of recommendation information to the user side, so that the user side stores the at least one piece of recommendation information, and updates the current recommendation result in the recommendation page at least according to the stored recommendation information.

The information recommendation apparatus shown in fig. 6 may execute the information recommendation method shown in the embodiment shown in fig. 2, and the implementation principle and the technical effect are not repeated. The detailed description of the information recommendation device in the above embodiments has been given in the embodiments related to the method, and the detailed description will not be given here.

In one possible design, the information recommendation apparatus of the embodiment shown in fig. 6 may be implemented as a server, which may include a storage group 701 and a processing component 702 as shown in fig. 7;

the storage component 701 stores one or more computer instructions for the processing component 702 to call and execute to implement the information recommendation method as shown in fig. 2.

Of course, the server may of course also comprise other components, such as input/output interfaces, communication components, etc.

The communication component is configured to facilitate wired or wireless communication between the server and other devices, and the like.

The server may be a cloud server in practical application, an elastic computing host provided for a cloud computing platform, and the like, and the processing component, the storage component, and the like may be basic server resources rented or purchased from the cloud computing platform.

An embodiment of the present application further provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a computer, the information recommendation method in the embodiment shown in fig. 2 may be implemented.

The processing components described in the respective embodiments above may include one or more processors executing computer instructions to perform all or part of the steps of the methods described above. Of course, the processing elements may also be implemented as one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, micro-controllers, microprocessors or other electronic components configured to perform the above-described methods.

The storage component is configured to store various types of data to support operations at the terminal. The memory components may be implemented by any type or combination of volatile or non-volatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.

The display element may be an Electroluminescent (EL) element, a liquid crystal display or a microdisplay having a similar structure, or a retina-directable display or similar laser scanning type display.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. Based on the understanding, the above technical solutions substantially or otherwise contributing to the prior art may be embodied in the form of a software product, which may be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method according to the various embodiments or some parts of the embodiments.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions in the embodiments of the present application.

Claims

1. An information recommendation method is applied to a user side, and the method comprises the following steps:

sending a first request to a server side, so that the server side determines at least one piece of recommendation information based on first user behavior data;

2. The method according to claim 1, wherein the server determines a first number of recommendation information based specifically on user behavior data;

the obtaining the at least one piece of recommendation information provided by the server and storing the at least one piece of recommendation information comprises:

acquiring the first quantity of recommendation information provided by the server;

sorting the first number of pieces of recommendation information and the stored recommendation information based on second user behavior data to obtain a first sorting result;

3. The method of claim 2, wherein updating the recommendation in the recommendation page based at least on the saved recommendation information comprises:

selecting a third amount of recommendation information from the stored recommendation information according to the first sequencing result;

and updating the current recommendation result in the recommendation page based on the third number of recommendation information.

4. The method of claim 3, further comprising: taking the unexposed quantity of the recommendation information in the current recommendation result as a third quantity;

the updating the recommendation result in the recommendation page based on the third number of recommendation information includes:

and updating the unexposed recommendation information in the recommendation page based on the third number of recommendation information.

5. The method of claim 1, wherein updating the current recommendation in the recommendation page based at least on the saved recommendation information comprises:

sorting the stored recommendation information and the recommendation information in the recommendation page based on third user behavior data to obtain a second sorting result;

and selecting a fourth amount of recommendation information to update the current recommendation result in the recommendation page according to the second sorting result.

6. The method of claim 1, wherein sending the first request to the server comprises:

determining at least one information type;

and sending a first request aiming at the at least one information type to a server, so that the server determines at least one piece of recommendation information corresponding to the at least one information type specifically based on the first user behavior data.

7. The method of claim 6, further comprising:

determining the number of requests respectively corresponding to the at least one information type;

the sending the first request for the at least one information type to the server comprises:

and sending a first request to a server based on the at least one information type and the request quantity respectively corresponding to the at least one information type, so that the server determines recommendation information with respective request quantity respectively corresponding to the at least one information type based on user behavior data.

8. The method of claim 1, wherein sending the first request to the server comprises:

and responding to a trigger event, and sending a first request to the server.

9. The method of claim 8, wherein the trigger event is generated when one or more of the following trigger event generation conditions are met:

detecting a user behavior satisfying a first specific condition;

detecting a paging trigger event;

the current time reaches the timing generation time of a preset period;

the stored recommendation information which accords with the user preference characteristics meets the triggering conditions;

detecting that the quantity of the stored recommendation information is smaller than a second quantity threshold value;

10. The method of claim 1, wherein updating the recommendation in the recommendation page based at least on the saved recommendation information comprises:

11. The method according to claim 10, wherein the update event is generated when one or more of the following update event generation conditions are satisfied:

detecting a user behavior satisfying a second specific condition;

identifying that a user has an updating intention;

12. The method of claim 1, further comprising:

sending a paging request to a server, so that the server responds to the paging request to generate a recommendation result;

acquiring the recommendation result provided by the server;

and displaying the recommendation result in the recommendation page.

13. The method of claim 12, wherein sending a paging request to a server comprises:

determining one or more information types;

and sending a paging request aiming at the one or more information types to a server.

14. The method of claim 1, wherein the current recommendation result is generated by the server based on a latest paging request or obtained after a latest update of the recommendation result shown in the recommendation page is performed at least according to the stored recommendation information.

15. The method of claim 1, wherein the recommendation information is merchandise related information.

16. An information recommendation method is applied to a server side, and the method comprises the following steps:

receiving a first request sent by a user side;

determining at least one recommendation based on the first user behavior data;

17. An information recommendation apparatus, comprising:

18. A user side is characterized by providing a display interface and a storage module;

and the display interface displays the current recommendation result, and updates the current recommendation result at least according to the recommendation information stored in the storage module.

19. An information recommendation apparatus, comprising:

20. An electronic device, comprising a display component, a storage component and a processing component;

the storage component stores one or more computer instructions; the one or more computer instructions for execution and invocation by a processing component to implement the information recommendation method of any one of claims 1-15.

21. A server is characterized by comprising a storage component and a processing component;

the storage component stores one or more computer instructions; the one or more computer instructions for execution and invocation by a processing component to implement the information recommendation method of claim 16.