CN114637927B - Content recommendation method and device, electronic equipment and storage medium - Google Patents
Content recommendation method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN114637927B CN114637927B CN202210497195.XA CN202210497195A CN114637927B CN 114637927 B CN114637927 B CN 114637927B CN 202210497195 A CN202210497195 A CN 202210497195A CN 114637927 B CN114637927 B CN 114637927B
- Authority
- CN
- China
- Prior art keywords
- content
- sequence
- target
- resource utilization
- service
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 44
- 230000006399 behavior Effects 0.000 claims description 12
- 238000010586 diagram Methods 0.000 description 11
- 230000002411 adverse Effects 0.000 description 6
- 230000000694 effects Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000004590 computer program Methods 0.000 description 4
- 238000011156 evaluation Methods 0.000 description 4
- 230000008707 rearrangement Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000001737 promoting effect Effects 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 101100190462 Caenorhabditis elegans pid-1 gene Proteins 0.000 description 1
- 101100190464 Caenorhabditis elegans pid-2 gene Proteins 0.000 description 1
- 101100190466 Caenorhabditis elegans pid-3 gene Proteins 0.000 description 1
- 101100243942 Caenorhabditis elegans pid-4 gene Proteins 0.000 description 1
- 101100243943 Caenorhabditis elegans pid-5 gene Proteins 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9538—Presentation of query results
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The present disclosure relates to a content recommendation method, apparatus, electronic device, and storage medium, the method comprising: acquiring a plurality of service content sequences, wherein each service content sequence corresponds to different service types, and each service content sequence comprises a plurality of service contents which are ordered based on the degree of correlation with a target object; performing sequence recall on the plurality of service content sequences based on preset sequence constraint information to obtain a plurality of mixed content sequences; determining the target resource utilization degree corresponding to each mixed content sequence; the target resource utilization degree represents the utilization degree of the corresponding mixed content sequence on the display position resource; selecting a target mixed content sequence from the plurality of mixed content sequences based on the target resource utilization degree corresponding to each mixed content sequence; and the target mixed content sequence is used for recommending to the target object. The present disclosure enables maximum utilization of exhibition location resources.
Description
Technical Field
The present disclosure relates to the field of computer technologies, and in particular, to a content recommendation method and apparatus, an electronic device, and a storage medium.
Background
The shuffling stage in the information stream recommendation scene is located after the fine shuffling and rearrangement stage of each service, and is used for performing mixed sorting processing on the content sequences from different services before issuing the final content sequence to the object, that is, the shuffling stage needs to process the service content sequences of different service types, for example, short videos uploaded by the object belong to a native service type, and advertisements belong to a promotion service type.
The mixed arrangement scheme in the related technology scores each service content entering a mixed arrangement stage based on a pointwise sequencing mechanism, and then exposes the service content from high to low according to the score of each service content, and the exposure content determined in the way cannot realize the maximum utilization of the display position resource.
Disclosure of Invention
The present disclosure provides a content recommendation method, apparatus, electronic device, and storage medium, to at least solve a problem in the related art that maximum utilization of display location resources cannot be achieved. The technical scheme of the disclosure is as follows:
according to a first aspect of an embodiment of the present disclosure, there is provided a content recommendation method including:
acquiring a plurality of service content sequences, wherein each service content sequence corresponds to different service types, and each service content sequence comprises a plurality of service contents which are ordered based on the degree of correlation with a target object;
performing sequence recall on the plurality of service content sequences based on preset sequence constraint information to obtain a plurality of mixed content sequences;
determining the target resource utilization degree corresponding to each mixed content sequence; the target resource utilization degree represents the utilization degree of the corresponding mixed content sequence on the display position resource;
selecting a target mixed content sequence from the plurality of mixed content sequences based on the target resource utilization degree corresponding to each mixed content sequence; and the target mixed content sequence is used for recommending to the target object.
In an exemplary embodiment, the determining the target resource utilization degree corresponding to each mixed content sequence includes:
determining the original resource utilization degree of the mixed content sequence based on the resource utilization degree of each service content in the mixed content sequence;
determining a target resource utilization loss of the mixed content sequence based on the resource utilization loss of the target service content in the mixed content sequence; the target service content is the service content corresponding to the service type constrained by the preset sequence constraint information;
and obtaining the target resource utilization degree of the mixed content sequence based on the original resource utilization degree and the target resource utilization loss of the mixed content sequence.
In an exemplary embodiment, the determining the original resource utilization degree of the mixed content sequence based on the resource utilization degree of each service content in the mixed content sequence includes:
acquiring the resource utilization degree of each service content in the mixed content sequence;
determining the position weight of each service content in the mixed content sequence; the position weight and the sequence position of the corresponding service content in the mixed content sequence form a positive correlation;
and carrying out weighted summation on the position weight of each service content in the mixed content sequence and the resource utilization degree of the corresponding service content to obtain the original resource utilization degree of the mixed content sequence.
In an exemplary embodiment, the obtaining the resource utilization degree of each service content in the mixed content sequence includes:
acquiring the resource utilization degree of each target service content in the mixed content sequence; the resource utilization degree of the target business content is determined according to the unit virtual resource consumption of the target business content and the object relevance degree, and the object relevance degree represents the relevance degree of the target business content and the target object;
acquiring the resource utilization degree of each non-target service content in the mixed content sequence; and determining the resource utilization degree of the non-target service content according to the historical operation behavior information corresponding to the non-target service content.
In an exemplary embodiment, the determining the position weight of each service content in the shuffled content sequence includes:
determining the sequence position of each target service content in the mixed content sequence and the sequence position of each non-target service content in the mixed content sequence;
determining a position weight corresponding to each target service content based on the sequence position of each target service content in the mixed content sequence;
and determining the position weight corresponding to each non-target service content based on the sequence position of each non-target service content in the mixed content sequence.
In an exemplary embodiment, the determining the target resource utilization loss of the shuffled content sequence based on the resource utilization loss of the target business content in the shuffled content sequence includes:
determining a content interval corresponding to each target service content in the mixed content sequence; the content interval refers to the number of service contents spaced between the corresponding target service content and the target service content ordered at the previous position;
determining resource utilization loss corresponding to each target service content based on the comparison condition of the content interval corresponding to each target service content in the mixed content sequence and an interval threshold;
and obtaining the target resource utilization loss of the mixed content sequence based on the resource utilization loss corresponding to each target service content in the mixed content sequence.
In an exemplary embodiment, the obtaining the target resource utilization loss of the mixed content sequence based on the resource utilization loss corresponding to each target service content in the mixed content sequence includes:
obtaining a first resource utilization loss corresponding to the mixed content sequence based on the resource utilization loss of the target service content in the mixed content sequence;
determining the highest sequence position of the service content corresponding to the constrained service type in the next recommendation based on the sequence position of the target service content in the mixed content sequence and the preset sequence constraint information;
determining a second resource utilization loss corresponding to the mixed content sequence based on the highest sequence position;
and obtaining the target resource utilization loss of the mixed content sequence based on the first resource utilization loss and the second resource utilization loss corresponding to the mixed content sequence.
In an exemplary embodiment, the determining, based on the highest sequence position, a second resource utilization loss corresponding to the shuffled content sequence includes:
determining the weight of the target position according to the highest sequence position;
and determining the second resource utilization loss corresponding to the mixed content sequence according to the target position weight.
In an exemplary embodiment, after the obtaining the plurality of service content sequences, the method further includes:
and determining the resource utilization degree of each target business content according to the unit virtual resource consumption and the object correlation degree corresponding to each target business content in the target business content sequence.
And determining the resource utilization degree of each non-target service content according to the historical operation behavior information corresponding to each non-target service content in the non-target service content sequence.
In an exemplary embodiment, an arrangement order of the service content of each service type in each of the shuffled content sequences in the shuffled content sequence is the same as an arrangement order of the service content in the service content sequence corresponding to the service type.
According to a second aspect of an embodiment of the present disclosure, there is provided a content recommendation apparatus including:
a content sequence acquiring unit configured to perform acquiring a plurality of service content sequences, each of the service content sequences corresponding to a different service type, each of the service content sequences including a plurality of service contents ordered based on a degree of correlation with a target object;
the sequence recall unit is configured to perform sequence recall on the plurality of service content sequences based on preset sequence constraint information to obtain a plurality of mixed content sequences;
a target resource utilization degree determination unit configured to perform determination of a target resource utilization degree corresponding to each of the shuffled content sequences; the target resource utilization degree represents the utilization degree of the corresponding mixed content sequence on the display position resource;
a target sequence selecting unit configured to select a target mixed content sequence from the plurality of mixed content sequences based on a target resource utilization degree corresponding to each mixed content sequence; and the target mixed content sequence is used for recommending to the target object.
In an exemplary embodiment, the target resource utilization determining unit includes:
an original resource utilization degree determining unit configured to determine original resource utilization degrees of the mixed content sequence based on resource utilization degrees of the business contents in the mixed content sequence;
a target resource utilization loss determination unit configured to perform determining a target resource utilization loss of the shuffled content sequence based on a resource utilization loss of a target business content in the shuffled content sequence; the target service content is the service content corresponding to the service type constrained by the preset sequence constraint information;
a target resource utilization degree determining subunit configured to perform obtaining a target resource utilization degree of the mixed content sequence based on the original resource utilization degree and the target resource utilization loss of the mixed content sequence.
In an exemplary embodiment, the original resource utilization determining unit includes:
a content resource utilization degree obtaining unit configured to perform obtaining resource utilization degrees of each service content in the mixed content sequence;
a position weight determination unit configured to perform determination of a position weight of each service content in the mixed content sequence; the position weight and the sequence position of the corresponding service content in the mixed content sequence form a positive correlation;
and the weighting unit is configured to perform weighted summation on the position weight of each service content in the mixed content sequence and the resource utilization degree of the corresponding service content to obtain the original resource utilization degree of the mixed content sequence.
In an exemplary embodiment, the content resource availability acquiring unit includes:
a first obtaining unit configured to perform obtaining resource utilization of each of the target service contents in the mixed content sequence; the resource utilization degree of the target business content is determined according to the unit virtual resource consumption of the target business content and the object correlation degree, and the object correlation degree represents the correlation degree of the target business content and the target object;
a second obtaining unit configured to perform obtaining resource utilization degrees of each non-target service content in the mixed content sequence; and determining the resource utilization degree of the non-target service content according to the historical operation behavior information corresponding to the non-target service content.
In an exemplary embodiment, the location weight determining unit includes:
a sequence position determining unit configured to perform determining a sequence position of each target service content in the mixed content sequence and a sequence position of each non-target service content in the mixed content sequence;
a first position weight determining subunit, configured to determine, based on the sequence position of each target service content in the mixed content sequence, a position weight corresponding to each target service content;
a second position weight determining subunit, configured to determine, based on the sequence position of each non-target service content in the mixed content sequence, a position weight corresponding to each non-target service content.
In an exemplary embodiment, the target resource utilization loss determining unit includes:
a content interval determining unit configured to perform determining a content interval corresponding to each target service content in the mixed content sequence; the content interval refers to the number of service contents spaced between the corresponding target service content and the target service content ordered at the previous position;
a content resource utilization loss determining unit configured to perform a comparison between a content interval corresponding to each target service content in the mixed content sequence and an interval threshold value, and determine a resource utilization loss corresponding to each target service content;
and the first determining unit is configured to execute resource utilization loss corresponding to each target service content in the mixed content sequence to obtain the target resource utilization loss of the mixed content sequence.
In an exemplary embodiment, the first determining unit includes:
a first resource utilization loss determining unit, configured to execute resource utilization loss based on target service content in the mixed content sequence, to obtain a first resource utilization loss corresponding to the mixed content sequence;
the highest sequence position determining unit is configured to determine the highest sequence position of the service content corresponding to the constrained service type in next recommendation based on the sequence position of the target service content in the mixed content sequence and the preset sequence constraint information;
a second resource usage loss determination unit configured to perform determining a second resource usage loss corresponding to the shuffled content sequence based on the highest sequence position;
a second determining unit configured to perform obtaining a target resource utilization loss of the shuffled content sequence based on a first resource utilization loss and a second resource utilization loss corresponding to the shuffled content sequence.
In an exemplary embodiment, the second resource utilization loss determining unit includes:
a third determining unit configured to perform determining a target location weight according to the highest sequence location;
a fourth determining unit configured to determine a second resource utilization loss corresponding to the shuffled content sequence according to the target location weight.
In an exemplary embodiment, the apparatus further comprises:
and the fifth determining unit is configured to determine the resource utilization degree of each target business content according to the unit virtual resource consumption and the object correlation degree corresponding to each target business content in the target business content sequence.
And the sixth determining unit is configured to determine the resource utilization degree of each non-target service content according to the historical operation behavior information corresponding to each non-target service content in the non-target service content sequence.
In an exemplary embodiment, an arrangement order of the service content of each service type in each of the shuffled content sequences in the shuffled content sequence is the same as an arrangement order of the service content in the service content sequence corresponding to the service type.
According to a third aspect of an embodiment of the present disclosure, there is provided an electronic apparatus including:
a processor;
a memory for storing the processor-executable instructions;
wherein the processor is configured to execute the instructions to implement the content recommendation method of the first aspect.
According to a fourth aspect of embodiments of the present disclosure, there is provided a computer-readable storage medium, wherein instructions, when executed by a processor of an electronic device, enable the electronic device to perform the content recommendation method of the first aspect.
According to a fifth aspect of embodiments of the present disclosure, there is provided a computer program product comprising computer programs/instructions which, when executed by a processor, implement the content recommendation method of the first aspect described above.
The technical scheme provided by the embodiment of the disclosure at least brings the following beneficial effects:
according to the method and the device, the multiple mixed-row content sequences are obtained by performing sequence recall on the multiple service content sequences of different service types based on the preset sequence constraint information, the target resource utilization degree corresponding to each mixed-row content sequence is further determined, and the target mixed-row content sequence is selected from the recalled multiple mixed-row content sequences to be recommended to a target object based on the target resource utilization degree corresponding to each mixed-row content sequence.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the disclosure and are not to be construed as limiting the disclosure.
FIG. 1 is a diagram illustrating an application environment for a method of content recommendation, according to an exemplary embodiment;
FIG. 2 is a flow diagram illustrating a method of content recommendation, according to an example embodiment;
FIG. 3 is a schematic flow diagram illustrating a process for determining a target resource utilization level for each shuffled content sequence in accordance with an exemplary embodiment;
FIG. 4 is an exemplary diagram illustrating the impact of exposure position on resource utilization of business content in accordance with one illustrative embodiment;
FIG. 5 is an illustration of an example of the impact of exposure density of targeted business content on exposure location resource utilization, in accordance with an illustrative embodiment;
FIG. 6 illustrates an example of the effect of a current shuffled content sequence on a subsequent sequence in accordance with an exemplary embodiment;
FIG. 7 is a schematic diagram of a shuffling technology architecture, shown in accordance with an exemplary embodiment;
FIG. 8 is a block diagram illustrating a content recommendation device in accordance with an exemplary embodiment;
FIG. 9 is a block diagram illustrating an electronic device in accordance with an example embodiment.
Detailed Description
In order to make the technical solutions of the present disclosure better understood, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.
It should be noted that the terms "first," "second," and the like in the description and claims of the present disclosure and in the above-described drawings are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the disclosure described herein are capable of operation in sequences other than those illustrated or otherwise described herein. The implementations described in the exemplary embodiments below are not intended to represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the disclosure, as detailed in the appended claims.
It should be further noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for presentation, analyzed data, etc.) referred to in the present disclosure are information and data authorized by the user or sufficiently authorized by each party.
Referring to fig. 1, a schematic diagram of an application environment of a content recommendation method according to an exemplary embodiment is shown, where the application environment may include a terminal 110 and a server 120, and the terminal 110 and the server 120 may be connected through a wired network or a wireless network.
The terminal 110 may be, but is not limited to, a smart phone, a tablet computer, a notebook computer, a desktop computer, and the like. The terminal 110 may have client software, such as an Application (App for short), installed therein, where the client software provides a service content display function, and the Application may be a stand-alone Application or a sub-program in the Application. Illustratively, the application may include a short video application, a live application, a shopping application, and the like. The user of the terminal 110 may log into the application through pre-registered user information, which may include an account number and a password.
The server 120 may be a server providing a background service for an application in the terminal 110, a content recommendation system is deployed on the server, the server 120 obtains a content recommendation request of a target object through the content recommendation system, performs mixed arrangement processing on a service content sequence generated by each service in response to the content recommendation request to obtain a mixed arrangement content sequence recommended to the target object, and sends the recommended mixed arrangement content sequence to the terminal 110 of the target object, where each service is a service with a different service type, such as a native service (live broadcast, short video, cold start, and the like), a promotion service (information flow service, payment promotion service, and the like), and the mixed arrangement content sequence recommended to the target object may include service contents with different service types.
The server 120 may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a network service, cloud communication, middleware service, a domain name service, a security service, a CDN, a big data and artificial intelligence platform, and the like.
Fig. 2 is a flowchart illustrating a content recommendation method according to an exemplary embodiment, and as shown in fig. 2, the content recommendation method is applied to the server in fig. 1, and includes the following steps.
In step S201, a plurality of service content sequences are obtained, each of the service content sequences corresponds to a different service type, and each of the service content sequences includes a plurality of service contents ordered based on a degree of correlation with a target object.
The service content sequence of each service type can be obtained by content recall, fine-ranking, coarse-ranking and rearrangement screening in sequence, generally, the service contents in each service content sequence are ranked from high to low according to the degree of correlation with the target object, and the higher the degree of correlation is, the more interested the target object is in the corresponding service contents, the more likely the target object is to browse the service contents. The business content may include short videos, live broadcasts, news, merchandise, and so on.
In a specific implementation, when a content recommendation request is initiated by a target object, service services call back service content from respective service content sets in response to the content recommendation request, and perform fine-arranging, coarse-arranging and rearrangement processing on the called-back service content in sequence to form a service content sequence of each service, and each service sends the respective service content sequence to a mixed arrangement service, so that the mixed arrangement service obtains the service content sequence of each service type.
In step S203, a plurality of mixed content sequences are obtained by performing sequence recall on the plurality of service content sequences based on preset sequence constraint information.
Wherein each of the shuffled content sequences includes service content of at least one service type.
The preset sequence constraint information is used for constraining the service contents of some service types in the mixed content sequence so as to ensure that the mixed content sequence recalled by the sequence meets some basic service requirements.
In practical application, the preset sequence constraint information may consider introducing control on exposure positions and exposure intervals of service contents of certain service types, for example, the preset sequence constraint information may be constraints on exposure positions and exposure intervals of promotion contents (such as advertisements) in each mixed content sequence, such as "there need to be 4 platform native contents at least between two promotion contents", "the first 3 exposure contents after a user enters a platform cannot be promotion contents", and the like, where the platform native contents may be contents uploaded to the platform by the user, and for example, the native contents in a short video platform may be recorded videos uploaded by the user, shot pictures, and the like.
In a specific implementation, the sequence recall may use an enumeration method to enumerate all mixed content sequences that satisfy the preset sequence constraint information.
In practical application, the ordering of each service content in the service content sequence entering the shuffling stage is an optimal decision optimized by the shuffling stage, and if the ordering sequence of each service content in the corresponding service content sequence is damaged in the shuffling stage, the final recommendation effect is adversely affected. Based on this, in an exemplary embodiment, in a plurality of mixed content sequences obtained based on the above-mentioned sequence recall, the service content of each service type in each mixed content sequence is arranged in the same order in the mixed content sequence as the service content sequence corresponding to the service type.
For example, a certain shuffled content sequence is { a1, B1, a2, a3, B2}, where the service contents a1, a2, a3 are from the service content sequence of service type a, and B1, B2 are from the service content sequence of service type B, then the order of the service contents a1, a2, a3 in the shuffled content sequence is the same as the order of the service contents in service type a, and the order of the service contents B1, B2 in the shuffled content sequence is the same as the order of the service contents in service type B.
In the embodiment, the sequence recall is performed on the premise of keeping the arrangement sequence of the service contents of each service type in the corresponding service content sequence, so that the optimal exposure sequence decided by each service in the rearrangement stage is prevented from being covered and damaged in the mixed arrangement stage, and the maximum utilization of the display position resource can be realized by the mixed arrangement content sequence screened subsequently.
In step S205, a target resource utilization degree corresponding to each of the mixed content sequences is determined.
The target resource utilization degree represents the utilization degree of the corresponding mixed content sequence on the display position resource, and is a value consideration of the whole sequence, and generally, the higher the utilization degree of the sequence on the display position resource is, the higher the sequence value of the corresponding mixed content sequence is.
In step S207, a target mixed content sequence is selected from the plurality of mixed content sequences based on a target resource utilization degree corresponding to each mixed content sequence.
Wherein the target shuffled content sequence is for recommendation to the target object. Generally, the utilization degree of the target resource corresponding to the target mixed content sequence is characterized by the maximum utilization degree of the display position resource. For example, when the target resource utilization degree is expressed in the form of a numerical value, the shuffled content sequence with the largest numerical value may be selected as the target shuffled content sequence.
In the embodiment, the utilization program of the display position resource by the whole sequence is considered from the perspective of the sequence, and the utilization degree of the display position resource by the whole sequence is actually integrated with the influence of the exposure position, so that the method is more suitable for an actual exposure scene, the display position resource can be maximally utilized, and the resource utilization rate is improved.
In an exemplary embodiment, the service content in the mixed content sequence includes a target service content and a non-target service content, where the target service content is a service content corresponding to a service type constrained by the preset sequence constraint information, and the non-target service content is a service content corresponding to a service type not constrained by the preset sequence constraint information, that is, the non-target service content is a service content in the mixed content sequence except the target service content. For example, if the preset sequence constraint information is that "there are at least 4 platform native contents needed between two promotion contents", the service type that the promotion contents are constrained may also be referred to as a target service type, and the service type that the platform native contents are unconstrained may also be referred to as a non-target service type. Therefore, the recalled mixed-ranking content sequence is constrained by the preset sequence constraint information in the sequence recall process, because the exposure of the constrained service type corresponding to the service content brings a certain loss of resource utilization of the display position.
In order to measure the utilization degree of target resources corresponding to the mixed content sequence, firstly, the resource utilization degree of the business content needs to be measured, the resource utilization degree of the business content represents the utilization degree of the business content on the display position resource, generally, the utilization degree of the business content on the display position resource can reflect the self value of the business content, the higher the utilization degree of the business content on the display position resource is, the higher the self value of the business content is, and conversely, the lower the utilization degree of the business content on the display position resource is, the lower the self value of the business content is.
For example, the value of the business content itself can be measured in terms of both the user dimension and the platform dimension, that is, the resource utilization of the business content can be measured in terms of both the user dimension and the platform dimension. Based on this, in step S201, after acquiring a plurality of service content sequences, the method further includes:
determining the resource utilization degree of each target business content according to the unit virtual resource consumption and the object correlation degree corresponding to each target business content in the target business content sequence; the object relevance degree characterizes the relevance degree of a corresponding target business object and the target object;
and determining the resource utilization degree of each non-target service content according to the historical operation behavior information corresponding to each non-target service content in the non-target service content sequence.
The target service content sequence is a service content sequence corresponding to a service type (namely, a target service type) constrained by preset sequence constraint information, and the non-target service content sequence is a service content sequence corresponding to a service type (namely, a non-target service type) which is not constrained.
Specifically, the historical operation behavior information corresponding to the non-target service content may include browsing duration, interaction rate, and the like, and generally, the longer the browsing duration is, the higher the interaction rate is, the higher the resource utilization degree of the non-target service content to the display position resource can be utilized to a greater extent by the non-target service content, that is, the higher the resource utilization degree of the non-target service content is. In order to improve the content recommendation efficiency, for example, the click-through rate pltr estimated by the non-target service content in the fine ranking stage may be obtained, and the resource utilization rate of the non-target service content may be determined based on the click-through rate pltr, so as to obtain the non-target service content in the non-target service content sequenceiFor example, the non-target service contentiThe resource utilization degree of (c) can be calculated by the following formula (1):
wherein,representing non-targeted business contentiResource utilization of (a);the weight corresponding to the non-target service type is represented and can be set according to actual needs;representing non-targeted business contentiCorresponding click-through rate.
For the target service content sequence, the target service content in the target service content sequence is usedjFor example, the target service contentjThe resource utilization degree of (c) can be calculated by the following formula (2):
whereinRepresenting targeted business contentjResource utilization of (a);andthe weight corresponding to the target service type is represented and can be set according to actual needs;representing object relevance;represents a unit of virtual resource consumption, such as the cost of consumption for an advertisement per thousand people presented. It should be noted that both the object correlation and the unit virtual resource consumption may be estimated based on the preamble link of the service corresponding to the target service content sequence.
According to the embodiment, the self values of the service contents of different service types are quantized from the two angles of the user dimension and the platform dimension respectively, so that the utilization degree of the display position resources of each service content can be accurately measured.
Based on this, in an exemplary embodiment, as shown in fig. 3, the step S205 may be implemented by:
in step S301, an original resource utilization degree of the mixed content sequence is determined based on the resource utilization degree of each service content in the mixed content sequence.
The original resource utilization degree of the mixed content sequence refers to the utilization degree of the display position resource of the mixed content sequence under the condition of not considering the resource utilization loss corresponding to the sequence. The resource utilization loss represents the negative influence on the resource utilization of the display position, the larger the resource utilization loss is, the larger the negative influence on the resource utilization of the display position is, and otherwise, the smaller the resource utilization loss is, the smaller the negative influence on the resource utilization of the display position is.
In the information flow scene, the utilization of the service content to the display position resource is embodied by exposure, and usually the user browses the content sequence from front to back, which results in that the probability of obtaining exposure for the content at the front of the sequence position is higher than that for the content at the back of the sequence position, that is, the exposure position will affect the self value of the service content in the mixed content sequence. In order to improve the accuracy of determining the target resource utilization rate corresponding to the mixed content sequence, the resource utilization rate of the service content needs to be corrected according to the sequence position of the service content in the mixed content sequence. In an exemplary embodiment, step S301, when implemented, may include:
acquiring the resource utilization degree of each service content in the mixed content sequence;
determining the position weight of each service content in the mixed content sequence; the position weight and the sequence position of the corresponding service content in the mixed content sequence form a positive correlation;
and carrying out weighted summation on the position weight of each service content in the mixed content sequence and the resource utilization degree of the corresponding service content to obtain the original resource utilization degree of the mixed content sequence.
The positive correlation relationship between the position weight and the sequence position of the corresponding service content in the mixed content sequence means that the position weight corresponding to the sequence position before is larger, and the position weight corresponding to the sequence position after is smaller.
In the above embodiment, the resource utilization of the service content is corrected by the sequence position of the service content in the mixed content sequence, which is beneficial to improving the accuracy of the subsequent determination of the original resource utilization of the mixed content sequence.
In an exemplary embodiment, obtaining the resource utilization degree of each service content in the mixed content sequence may include: and acquiring the resource utilization degree of each target business content in the mixed content sequence and the resource utilization degree of each non-target business content in the mixed content sequence. By respectively obtaining the resource utilization degrees of the target service content and the non-target service content in the mixed content sequence, the original resource utilization degree of the mixed content sequence can be accurately obtained.
Accordingly, in an exemplary embodiment, determining the location rights of each business content in the shuffled content sequence may include:
determining the sequence position of each target service content in the mixed content sequence and the sequence position of each non-target service content in the mixed content sequence;
determining a position weight corresponding to each target service content based on the sequence position of each target service content in the mixed content sequence;
and determining the position weight corresponding to each non-target service content based on the sequence position of each non-target service content in the mixed content sequence.
Specifically, when determining the position weight of each service content in the mixed content sequence, the target service content and the non-target service content in the mixed content sequence are considered respectively, so that the influence of the exposure position on the self value of the service contents of different service types can be measured more accurately, and the resource utilization rate of each service content in the mixed content sequence can be corrected more accurately.
In particular embodiments, the location weight of the non-targeted business contentAnd location weighting of target business contentCan be calculated by the following equations (3) and (4):
wherein,;andare all taken as values of [0, 1%]Parameter values in between, for example both may take 0.9;andthe position numbers in the shuffled content sequence are usually consecutive arabic numerals starting from 0, that is, the position number of the first sequence is 0, and then sequentially increases toward the position number of the last sequence. The position weight can show a change trend from big to small from the first position of the sequence to the end of the sequence through the calculation formula.
Taking non-target service content as native content (such as short video uploaded by a user, live broadcast, etc.) and target service content as promotion content (such as advertisement), the calculation of the position weight corresponding to the target service content will be exemplarily described below with reference to fig. 4. As shown in FIG. 4, the promotion content ad1 in the mixed content sequence 1 is located at the first position of the sequence and the position serial numberposIf the position weight is 0, the position weight corresponding to the promotion information ad1 in the mixed content sequence 1 is(ii) a The promotion information ad1 in the mixed content sequence 2 is positioned at the 2 nd position of the sequence and has a position serial numberposIf the position weight is 1, the position weight corresponding to the promotion information ad1 in the mixed content sequence 2 is。
It will be appreciated that the original resource utilization of the shuffled content sequence can be obtained by weighting and summing the location weights of the non-targeted traffic content in the shuffled content sequence and the resource utilization of the corresponding non-targeted traffic contentAnd carrying out weighted summation based on the position weight of each target service content in the mixed content sequence and the resource utilization degree of the corresponding target service content to obtainThen combining the two weighted summation results to obtain the original resource utilization rate of the mixed content sequence, i.e. the mixed content sequenceThe original resource utilization of the capacity sequence can be expressed as the following equation (5):
wherein,representing non-targeted traffic content in a shuffled content sequenceiPosition weight of [ 1, 0 ], which is a value of]In the middle of;representing non-targeted business contentiResource utilization of.
wherein,representing targeted business content in a shuffled content sequencejPosition weight of [ 1, 0 ], which is a value of]To (c) to (d);representing targeted business contentjResource utilization of.
In step S303, a target resource utilization loss of the shuffled content sequence is determined based on the resource utilization loss of the target service content in the shuffled content sequence.
Because the utilization of the display position resource of the sequence by the target service content has a certain adverse effect, especially when the target service content in the mixed content sequence is dense, the adverse effect is larger, and on the premise that the preset sequence constraint information is satisfied, the mixed content sequence with larger target service content interval can better utilize the display position resource compared with the mixed content sequence with smaller target service content interval. Therefore, when measuring the target resource utilization of the mixed content sequence, the adverse effect of the exposure density of the target service content in the mixed content sequence on the display position resource utilization needs to be considered, so as to improve the accuracy of the target resource utilization corresponding to the mixed content sequence.
Based on this, in order to improve the accuracy of recommendation based on the target resource utilization of the shuffled content sequence, in an exemplary embodiment, the step S303 may include:
determining a content interval corresponding to each target service content in the mixed content sequence;
determining resource utilization loss corresponding to each target service content based on the comparison condition of the content interval corresponding to each target service content in the mixed content sequence and a content interval threshold;
and obtaining the target resource utilization loss of the mixed content sequence based on the resource utilization loss corresponding to each target service content in the mixed content sequence.
Wherein, the content interval refers to the number of the service contents spaced between the corresponding target service content and the target service content ranked at the previous position. It should be noted that, the target service content ordered at the previous position may be the target service content ordered at the previous position in the mixed content sequence where the target service content is located; if the target service content is the first target service content to be exposed in the mixed content sequence, the target service content ordered in the previous position is the last history target service content exposed in the history mixed content sequence exposed last time.
As shown in fig. 5, in the current mixed content sequence (which is any one of the mixed content sequences), the target traffic content ordered one bit before the target traffic content ad1 is the historical target traffic content ad in the history mixed content sequence exposed most recently, the target traffic content ordered one bit before the target traffic content ad2 is ad1, the content interval corresponding to ad1 is 4, and the content interval corresponding to ad2 is 5.
In the embodiment of the present disclosure, when the content interval corresponding to the target service content is smaller than the content interval threshold, it is considered that the content interval corresponding to the target service content does not adversely affect the utilization of the display location resource.
In a specific implementation, the resource utilization loss corresponding to each target service content in the mixed content sequence can be calculated by the following formula (8):
wherein,representing targeted business content in a shuffled content sequencejResource utilization loss of (2);representing targeted business content in a shuffled content sequencejThe content interval of (a);a presentation content interval threshold;is a preset parameter value with the value of [0, 1%]For example, 0.5 may be used.
For example, the sum of the resource utilization losses corresponding to the target service contents in the shuffled content sequence may be used as the target resource utilization loss of the shuffled content sequence.
In the above embodiment, the adverse effect of the exposure density of the target service content in the mixed content sequence on the resource utilization of the display position is measured, which is beneficial to improving the accuracy of the utilization rate of the target resource corresponding to the mixed content sequence, and further improving the accuracy of content recommendation.
In step S305, a target resource utilization degree of the mixed content sequence is obtained based on the original resource utilization degree and the target resource utilization loss of the mixed content sequence.
Specifically, the target resource utilization of the shuffled content sequence may be determined from a difference between an original resource utilization of the shuffled content sequence and a target resource utilization loss of the shuffled content sequence.
In the above embodiment, the utilization condition of each mixed content sequence on the display position resource is evaluated from the sequence dimension in combination with the resource utilization degree and the resource utilization loss of the sequence, and the largest target mixed content sequence is selected as the target mixed content sequence to be recommended to the user based on the target resource utilization degree obtained by the evaluation, so that the maximum utilization of the display position resource can be realized.
In consideration of the fact that in practical application, a user can continuously request a next mixed content sequence in the process of browsing information, as a part of context, a historical exposure sequence can influence the decision of a subsequent sequence. Therefore, in the decision process of the shuffling stage, the value maximization of the shuffling sequence in a single request cannot be considered, and the influence of the current sequence on the value of the subsequent sequence should be considered, so as to obtain the shuffling scheme for globally showing the position resource utilization maximization in a use time period. The following describes the influence of the current sequence on the subsequent sequence in a specific application scenario.
Assume that a minimum of 4 pieces of native content (e.g., short video recorded by a user, live broadcast, etc.) are defined between two promotional content (e.g., advertisements) during a sequence recall. For the current request, two pieces of promotion content are to be released, and 1 piece of promotion content, 2 pieces of promotion content or no promotion content can be selectively released, wherein the content resource utilization information corresponding to the promotion content ad2Is a very small positive number, i.e., the promotional content has a low utilization of the exhibition location resource. However, if only the value maximization of the sequence in the current request is considered, the value maximization is shown in FIG. 6For the mixed content sequence 1 and the mixed content sequence 2 recalled by the previous request sequence, because one more promotion content is released in the mixed content sequence 1, the sequence resource utilization information of the mixed content sequence 1 is higher than that of the sequence resource utilization information of the sequence 2. And if the influence of the current sequence on the subsequent sequence is taken into consideration, the promotion content ad is arranged in the next exposure sequence corresponding to the mixed content sequence 2 under the limit that the promotion content interval is not less than 4 next The highest energy is arranged at the first of the sequence, and the promotion content ad is arranged in the next exposure sequence corresponding to the mixed content sequence 1 next The highest content can only be arranged at the 5 th position of the sequence, and the popularization content ad can be known by combining the formulas (4) and (7) next Can be affected by a significant positional compromise. Therefore, after considering the influence of the current sequence on the subsequent sequence, the shuffled content sequence 1 is not necessarily better than the shuffled content sequence 2 from the viewpoint of the maximum utilization of the global presentation position resource.
Based on this, in an exemplary embodiment, the obtaining the target resource utilization loss of the shuffled content sequence based on the resource utilization loss corresponding to each target service content in the shuffled content sequence may include:
obtaining a first resource utilization loss corresponding to the mixed content sequence based on the resource utilization loss of the target service content in the mixed content sequence;
determining the highest sequence position of the service content corresponding to the constrained service type in the next recommendation based on the sequence position of the target service content in the mixed content sequence and the preset sequence constraint information;
determining a second resource utilization loss corresponding to the mixed content sequence based on the highest sequence position;
and obtaining the target resource utilization loss of the mixed content sequence based on the first resource utilization loss and the second resource utilization loss corresponding to the mixed content sequence.
Specifically, the first resource utilization loss for the shuffled content sequence may be calculated based on the following equation (9):
wherein,a first resource utilization penalty representing a shuffled content sequence;represents the loss weight of the first sequence and takes the value of [0, 1%]Specific numerical values can be set according to actual experience;representing targeted business content in a shuffled content sequencejThe content resource utilization loss.
Taking the current shuffle content sequence shown in FIG. 5 as an example, the settings are set Then the first resource utilization penalty of the current shuffled content sequence shown in fig. 5 is。
In the above embodiment, from the angle of the influence of the current mixed content sequence on resource utilization of the subsequent sequence display position, the highest sequence position of the service content of the target service type in the next recommendation is determined based on the sequence position of the target service content in the mixed content sequence and the preset sequence constraint information, and the second resource utilization loss of the mixed content sequence is determined based on the highest sequence position, so that the target resource utilization loss of the mixed content sequence is determined by combining the first resource utilization loss and the second resource utilization loss, that is, the target resource utilization loss of the mixed content sequence is formed by the first resource utilization loss and the second resource utilization loss of the mixed content sequence, which is favorable for realizing maximization of resource utilization of the global display position within a period of time. It should be noted that, in the embodiments of the present disclosure, the sequence position is higher before the sequence position is closer, that is, the first sequence is the highest sequence position, the last sequence is the lowest sequence position, and the sequence positions between the first sequence and the last sequence are sequentially lower.
Specifically, the sequence position of the last exposed target service content in the mixed content sequence may be determined, then the number of the minimum service contents requiring the interval between adjacent target service contents is determined based on the preset sequence constraint information, and the highest sequence position of the service content of the target service type in the next recommendation may be determined based on the number of the minimum service contents at the interval and the sequence position. Taking the mixed content sequence 1 in fig. 6 as an example, the target service content exposed last is located at the end of the sequence, and assuming that the preset sequence constraint information is that at least 4 pieces of non-target service content are spaced between two target service contents, it may be determined that the highest sequence position of the service content of the target service type in the next recommendation is 5.
In an exemplary embodiment, when determining the second resource utilization loss of the shuffled content sequence based on the above-described highest-sequence position may include:
determining the weight of the target position according to the highest sequence position;
and determining the second resource utilization loss of the mixed content sequence according to the target position weight.
In a specific implementation, the target position weight can be calculated based on the foregoing formula (4), that is, the target position weight can be expressed asWhereinand the position serial number of the highest sequence position corresponding to the service content of the target service type in the next recommendation. Then the second resource utilization loss of the shuffled content sequence can be calculated by the following equation (10)Calculating:
wherein,a second resource utilization penalty representing a shuffled content sequence;the weight indicating the second sequence loss may be set according to practical experience, or may be set based on the contribution value of the target object to the platform in practical application, for example, the higher the contribution value is, the higher the corresponding contribution value isThe larger may be.
The second resource utilization loss of the mixed content sequence can be obtained more accurately by combining the target position weight of the highest sequence position in the above embodiment.
Illustratively, the target resource utilization of the shuffled content sequence may be calculated based on the following equation (11):
In the above embodiment, when determining the target resource utilization degree corresponding to the mixed content sequence, the influence of the exposure position on the resource utilization degree of the business content, the influence of the exposure density of the target business content on the resource utilization degree of the display position, and the influence of the current mixed content sequence on the resource utilization degree of the display position of the subsequent sequence are comprehensively considered, so that the utilization condition of each mixed content sequence on the display position resource within a period of time is more comprehensively evaluated from the sequence dimension, the maximum mixed content sequence is selected as the target mixed content sequence to be recommended to the user based on the target resource utilization degree obtained by the evaluation, and the maximum utilization of the display position resource within a period of time is realized.
In order to better understand the technical solution of the embodiment of the present disclosure, the following description is made with reference to the mixed-arranging technical architecture diagram of fig. 7, taking non-target service content as native content, and taking a target service type as popularization content as an example.
As shown in fig. 7, the native content may include short videos recorded by users, live and cold start service content, and the like, and the promotion content may include information stream advertisements, paid promotion content, and the like. In the mixed ranking stage, acquiring a native content sequence { pid1, pid2, pid3, pid4, pid5 and pid6} and a promotion content sequence { ad1, ad2}, { fans1 and fans2 }; then, the first calculating module calculates, for each sequence, the resource utilization of each content in the sequence, which is presented in a form of a value in fig. 7, and the specific calculating manner may refer to the foregoing related description about the resource utilization of the service content in the embodiment of the present disclosure, and is not described herein again; then, the sequence recall module recalls the native content sequence and the popularization content sequence in sequence based on preset sequence constraint information to obtain a recall sequence set comprising a plurality of mixed content sequences; and then entering a sequence evaluation module, evaluating each mixed content sequence by the sequence evaluation module from three angles of influence of the exposure position on resource utilization of each content, influence of popularization content density on resource utilization of the display position and influence of the current sequence on resource utilization of the display position of a subsequent sequence to obtain target resource utilization of each mixed content sequence, presenting the target resource utilization in a sequence value form, and issuing the mixed content sequence with the maximum sequence value as an optimal mixed content sequence.
Fig. 8 is a block diagram illustrating a content recommendation device according to an example embodiment. Referring to fig. 8, the content recommendation apparatus 800 includes:
a content sequence obtaining unit 810 configured to perform obtaining a plurality of service content sequences, each of the service content sequences corresponding to a different service type, each of the service content sequences including a plurality of service contents ordered based on a degree of correlation with a target object;
a sequence recall unit 820 configured to perform a sequence recall on the plurality of service content sequences based on preset sequence constraint information to obtain a plurality of mixed content sequences;
a target resource utilization determining unit 830 configured to perform determining a target resource utilization corresponding to each of the shuffled content sequences; the target resource utilization degree represents the utilization degree of the corresponding mixed content sequence to the display position resource;
a target sequence selecting unit 840 configured to perform selecting a target mixed content sequence from the plurality of mixed content sequences based on a target resource utilization degree corresponding to each mixed content sequence; and the target mixed content sequence is used for recommending to the target object.
In an exemplary embodiment, the target resource utilization determining unit includes:
an original resource utilization degree determining unit configured to determine original resource utilization degrees of the mixed content sequence based on resource utilization degrees of the business contents in the mixed content sequence;
a target resource utilization loss determination unit configured to perform determining a target resource utilization loss of the shuffled content sequence based on a resource utilization loss of a target business content in the shuffled content sequence; the target service content is the service content corresponding to the service type constrained by the preset sequence constraint information;
a target resource utilization degree determining subunit configured to perform obtaining a target resource utilization degree of the mixed content sequence based on the original resource utilization degree and the target resource utilization loss of the mixed content sequence.
In an exemplary embodiment, the original resource utilization determining unit includes:
a content resource utilization degree obtaining unit configured to perform obtaining resource utilization degrees of each service content in the mixed content sequence;
a position weight determination unit configured to perform determination of a position weight of each service content in the mixed content sequence; the position weight and the sequence position of the corresponding service content in the mixed content sequence form a positive correlation;
and the weighting unit is configured to perform weighted summation on the position weight of each service content in the mixed content sequence and the resource utilization degree of the corresponding service content to obtain the original resource utilization degree of the mixed content sequence.
In an exemplary embodiment, the content resource utilization obtaining unit includes:
a first obtaining unit configured to perform obtaining resource utilization of each of the target service contents in the mixed content sequence; the resource utilization degree of the target business content is determined according to the unit virtual resource consumption of the target business content and the object relevance degree, and the object relevance degree represents the relevance degree of the target business content and the target object;
a second obtaining unit configured to perform obtaining resource utilization degrees of each non-target service content in the mixed content sequence; and determining the resource utilization degree of the non-target service content according to the historical operation behavior information corresponding to the non-target service content.
In an exemplary embodiment, the location weight determining unit includes:
a sequence position determining unit configured to perform determining a sequence position of each target service content in the mixed content sequence and a sequence position of each non-target service content in the mixed content sequence;
a first position weight determining subunit, configured to perform determining, based on sequence positions of the target business contents in the mixed content sequence, a position weight corresponding to each of the target business contents;
and the second position weight determining subunit is configured to determine the position weight corresponding to each non-target service content based on the sequence position of each non-target service content in the mixed content sequence.
In an exemplary embodiment, the target resource utilization loss determining unit includes:
a content interval determining unit configured to perform determining a content interval corresponding to each target service content in the mixed content sequence; the content interval refers to the number of the service contents between the corresponding target service content and the target service content sequenced at the previous position;
a content resource utilization loss determining unit configured to perform a comparison between a content interval corresponding to each target service content in the mixed content sequence and an interval threshold value, and determine a resource utilization loss corresponding to each target service content;
and the first determining unit is configured to execute resource utilization loss corresponding to each target service content in the mixed content sequence to obtain the target resource utilization loss of the mixed content sequence.
In an exemplary embodiment, the first determining unit includes:
a first resource utilization loss determining unit, configured to execute resource utilization loss based on target service content in the mixed content sequence, to obtain a first resource utilization loss corresponding to the mixed content sequence;
the highest sequence position determining unit is configured to determine the highest sequence position of the service content corresponding to the constrained service type in the next recommendation based on the sequence position of the target service content in the mixed content sequence and the preset sequence constraint information;
a second resource usage loss determination unit configured to perform determining a second resource usage loss corresponding to the shuffled content sequence based on the highest sequence position;
a second determining unit configured to perform obtaining a target resource utilization loss of the shuffled content sequence based on a first resource utilization loss and a second resource utilization loss corresponding to the shuffled content sequence.
In an exemplary embodiment, the second resource utilization loss determining unit includes:
a third determination unit configured to perform determining a target location weight according to the highest sequence location;
and the fourth determining unit is configured to determine a second resource utilization loss corresponding to the mixed content sequence according to the target position weight.
In an exemplary embodiment, the apparatus further comprises:
and the fifth determining unit is configured to determine the resource utilization degree of each target business content according to the unit virtual resource consumption amount and the object correlation degree corresponding to each target business content in the target business content sequence.
And the sixth determining unit is configured to determine the resource utilization degree of each non-target service content according to the historical operation behavior information corresponding to each non-target service content in the non-target service content sequence.
In an exemplary embodiment, an arrangement order of the service content of each service type in each of the shuffled content sequences in the shuffled content sequence is the same as an arrangement order of the service content in the service content sequence corresponding to the service type.
With regard to the apparatus in the above-described embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.
In one exemplary embodiment, there is also provided an electronic device, comprising a processor; a memory for storing processor-executable instructions; wherein the processor is configured to implement the content recommendation method provided in any of the above embodiments when executing the instructions stored in the memory.
The electronic device may be a terminal, a server, or a similar computing device, taking the electronic device as a server as an example, fig. 9 is a block diagram of an electronic device for content recommendation shown according to an exemplary embodiment, and as shown in fig. 9, the server 900 may have a relatively large difference due to different configurations or performances, and may include one or more Central Processing Units (CPUs) 910 (the processor 910 may include but is not limited to a Processing device such as a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 930 for storing data, and one or more storage media 920 (e.g., one or more mass storage devices) for storing an application 923 or data 922. Memory 930 and storage media 920 may be, among other things, transient or persistent storage. The program stored in the storage medium 920 may include one or more modules, each of which may include a series of instruction operations in a server. Still further, the central processor 910 may be configured to communicate with the storage medium 920, and execute a series of instruction operations in the storage medium 920 on the server 900. The server 900 may also include one or more power supplies 960, one or more wired or wireless network interfaces 950, one or more input-output interfaces 940, and/or one or more operating systems 921, such as Windows Server, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, and so forth.
The input/output interface 940 may be used to receive or transmit data via a network. Specific examples of the network described above may include a wireless network provided by a communication provider of the server 900. In one example, the input/output Interface 940 includes a Network adapter (NIC) that can be connected to other Network devices through a base station to communicate with the internet. In one example, the input/output interface 940 may be a Radio Frequency (RF) module, which is used for communicating with the internet in a wireless manner.
It will be understood by those skilled in the art that the structure shown in fig. 9 is only an illustration and is not intended to limit the structure of the electronic device. For example, server 900 may also include more or fewer components than shown in FIG. 9, or have a different configuration than shown in FIG. 9.
In an exemplary embodiment, a computer readable storage medium comprising instructions, such as the memory 930 comprising instructions, executable by the processor 910 of the apparatus 9000 to perform the above method is also provided. Alternatively, the computer readable storage medium may be a ROM, a Random Access Memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.
In an exemplary embodiment, there is also provided a computer program product comprising computer programs/instructions which, when executed by a processor, implement the content recommendation method provided in any of the above embodiments.
Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This application is intended to cover any variations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.
It will be understood that the present disclosure is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.
Claims (20)
1. A content recommendation method, comprising:
acquiring a plurality of service content sequences, wherein each service content sequence corresponds to different service types, and each service content sequence comprises a plurality of service contents which are ordered based on the degree of correlation with a target object;
performing sequence recall on the plurality of service content sequences based on preset sequence constraint information to obtain a plurality of mixed content sequences;
determining target resource utilization degree corresponding to each mixed content sequence based on the original resource utilization degree and the target resource utilization loss of each mixed content sequence; the target resource utilization degree represents the utilization degree of the corresponding mixed content sequence to the display position resource;
selecting a target mixed content sequence from the plurality of mixed content sequences based on the target resource utilization degree corresponding to each mixed content sequence; the target mixed content sequence is used for being recommended to the target object;
wherein the target resource utilization loss of the shuffled content sequence is obtained based on:
determining content intervals corresponding to all target service contents in the mixed content sequence; the content interval refers to the number of service contents spaced between corresponding target service contents and target service contents sequenced at the previous position, and the target service contents are service contents corresponding to the service types constrained by the preset sequence constraint information;
determining resource utilization loss corresponding to each target service content based on the comparison condition of the content interval corresponding to each target service content in the mixed content sequence and an interval threshold;
and obtaining the target resource utilization loss of the mixed content sequence based on the resource utilization loss corresponding to each target service content in the mixed content sequence.
2. The content recommendation method according to claim 1, further comprising: and determining the original resource utilization degree of the mixed content sequence based on the resource utilization degree of each service content in the mixed content sequence.
3. The content recommendation method according to claim 2, wherein said determining an original resource utilization degree of the shuffled content sequence based on a resource utilization degree of each business content in the shuffled content sequence comprises:
acquiring the resource utilization degree of each service content in the mixed content sequence;
determining the position weight of each service content in the mixed content sequence; the position weight and the sequence position of the corresponding service content in the mixed content sequence form a positive correlation;
and carrying out weighted summation on the position weight of each service content in the mixed content sequence and the resource utilization degree of the corresponding service content to obtain the original resource utilization degree of the mixed content sequence.
4. The content recommendation method according to claim 3, wherein said obtaining resource utilization of each service content in the mixed content sequence comprises:
acquiring the resource utilization degree of each target service content in the mixed content sequence; the resource utilization degree of the target business content is determined according to the unit virtual resource consumption of the target business content and the object relevance degree, and the object relevance degree represents the relevance degree of the target business content and the target object;
acquiring the resource utilization degree of each non-target service content in the mixed content sequence; and determining the resource utilization degree of the non-target service content according to the historical operation behavior information corresponding to the non-target service content.
5. The content recommendation method according to claim 3, wherein said determining the position weight of each service content in the shuffled content sequence comprises:
determining the sequence position of each target service content in the mixed content sequence and the sequence position of each non-target service content in the mixed content sequence;
determining a position weight corresponding to each target service content based on the sequence position of each target service content in the mixed content sequence;
and determining the position weight corresponding to each non-target service content based on the sequence position of each non-target service content in the mixed content sequence.
6. The content recommendation method according to claim 1, wherein the obtaining of the target resource utilization loss of the mixed content sequence based on the resource utilization loss corresponding to each target service content in the mixed content sequence comprises:
obtaining a first resource utilization loss corresponding to the mixed content sequence based on the resource utilization loss of the target service content in the mixed content sequence;
determining the highest sequence position of the service content corresponding to the constrained service type in the next recommendation based on the sequence position of the target service content in the mixed content sequence and the preset sequence constraint information;
determining a second resource utilization loss corresponding to the mixed content sequence based on the highest sequence position;
and obtaining the target resource utilization loss of the mixed content sequence based on the first resource utilization loss and the second resource utilization loss corresponding to the mixed content sequence.
7. The content recommendation method according to claim 6, wherein said determining a second resource utilization loss corresponding to the shuffled content sequence based on the highest sequence position comprises:
determining the weight of the target position according to the highest sequence position;
and determining the second resource utilization loss corresponding to the mixed content sequence according to the target position weight.
8. The content recommendation method according to claim 2, wherein after said obtaining a plurality of service content sequences, said method further comprises:
determining the resource utilization degree of each target business content according to the unit virtual resource consumption and the object correlation degree corresponding to each target business content in the target business content sequence;
and determining the resource utilization degree of each non-target service content according to the historical operation behavior information corresponding to each non-target service content in the non-target service content sequence.
9. The content recommendation method according to any one of claims 1 to 8, wherein an arrangement order of the service content of each service type in each of the shuffled content sequences in the shuffled content sequences is the same as an arrangement order of the service content sequences corresponding to the service type.
10. A content recommendation apparatus characterized by comprising:
a content sequence acquiring unit configured to perform acquiring a plurality of service content sequences, each of the service content sequences corresponding to a different service type, each of the service content sequences including a plurality of service contents ordered based on a degree of correlation with a target object;
the sequence recalling unit is configured to perform sequence recalling on the plurality of service content sequences based on preset sequence constraint information to obtain a plurality of mixed content sequences;
a target resource utilization degree determining unit configured to determine a target resource utilization degree corresponding to each mixed content sequence based on an original resource utilization degree and a target resource utilization loss of each mixed content sequence; the target resource utilization degree represents the utilization degree of the corresponding mixed content sequence on the display position resource;
a target sequence selecting unit configured to select a target mixed content sequence from the plurality of mixed content sequences based on a target resource utilization degree corresponding to each mixed content sequence; the target mixed content sequence is used for being recommended to the target object;
wherein the target resource utilization degree determination unit includes a target resource utilization loss determination unit including:
a content interval determining unit configured to perform determining a content interval corresponding to each target service content in the mixed content sequence; the content interval refers to the number of service contents spaced between corresponding target service contents and target service contents sequenced at the previous position, and the target service contents are service contents corresponding to the service types constrained by the preset sequence constraint information;
a content resource utilization loss determining unit configured to perform a comparison between a content interval corresponding to each target service content in the mixed content sequence and an interval threshold value, and determine a resource utilization loss corresponding to each target service content;
and the first determining unit is configured to execute resource utilization loss corresponding to each target service content in the mixed content sequence to obtain the target resource utilization loss of the mixed content sequence.
11. The content recommendation device according to claim 10, wherein the target resource utilization degree determination unit further comprises:
an original resource utilization degree determination unit configured to perform determining an original resource utilization degree of the mixed content sequence based on the resource utilization degree of each service content in the mixed content sequence.
12. The content recommendation device according to claim 11, wherein the original resource utilization degree determination unit comprises:
a content resource utilization degree obtaining unit configured to perform obtaining resource utilization degrees of each service content in the mixed content sequence;
a position weight determination unit configured to perform determination of a position weight of each business content in the shuffled content sequence; the position weight and the sequence position of the corresponding service content in the mixed content sequence form a positive correlation;
and the weighting unit is configured to perform weighted summation on the position weight of each service content in the mixed content sequence and the resource utilization degree of the corresponding service content to obtain the original resource utilization degree of the mixed content sequence.
13. The content recommendation device according to claim 12, wherein the content resource availability obtaining unit includes:
a first obtaining unit configured to perform obtaining resource utilization of each of the target service contents in the mixed content sequence; the resource utilization degree of the target business content is determined according to the unit virtual resource consumption of the target business content and the object correlation degree, and the object correlation degree represents the correlation degree of the target business content and the target object;
a second obtaining unit configured to perform obtaining resource utilization degrees of each non-target service content in the mixed content sequence; and determining the resource utilization degree of the non-target service content according to the historical operation behavior information corresponding to the non-target service content.
14. The content recommendation device according to claim 12, wherein the position weight determination unit includes:
a sequence position determining unit configured to perform determining a sequence position of each target service content in the mixed content sequence and a sequence position of each non-target service content in the mixed content sequence;
a first position weight determining subunit, configured to determine, based on the sequence position of each target service content in the mixed content sequence, a position weight corresponding to each target service content;
a second position weight determining subunit, configured to determine, based on the sequence position of each non-target service content in the mixed content sequence, a position weight corresponding to each non-target service content.
15. The content recommendation device according to claim 10, wherein the first determination unit includes:
a first resource utilization loss determining unit, configured to execute resource utilization loss based on target service content in the mixed content sequence, to obtain a first resource utilization loss corresponding to the mixed content sequence;
the highest sequence position determining unit is configured to determine the highest sequence position of the service content corresponding to the constrained service type in the next recommendation based on the sequence position of the target service content in the mixed content sequence and the preset sequence constraint information;
a second resource utilization loss determination unit configured to perform determination of a second resource utilization loss corresponding to the shuffled content sequence based on the highest sequence position;
a second determining unit configured to perform obtaining a target resource utilization loss of the shuffled content sequence based on a first resource utilization loss and a second resource utilization loss corresponding to the shuffled content sequence.
16. The content recommendation device according to claim 15, wherein the second resource utilization loss determination unit includes:
a third determination unit configured to perform determining a target location weight according to the highest sequence location;
and the fourth determining unit is configured to determine a second resource utilization loss corresponding to the mixed content sequence according to the target position weight.
17. The content recommendation device according to claim 11, wherein the device further comprises:
a fifth determining unit, configured to determine resource utilization of each target business content according to the unit virtual resource consumption and the object correlation corresponding to each target business content in the target business content sequence;
and the sixth determining unit is configured to execute determining the resource utilization degree of each non-target service content according to the historical operation behavior information corresponding to each non-target service content in the non-target service content sequence.
18. The content recommendation device according to any one of claims 10 to 17, wherein an arrangement order of the service content of each service type in each of the shuffled content sequences in the shuffled content sequences is the same as an arrangement order of the service content sequences corresponding to the service type.
19. An electronic device, comprising:
a processor;
a memory for storing the processor-executable instructions;
wherein the processor is configured to execute the instructions to implement the content recommendation method of any one of claims 1 to 9.
20. A computer-readable storage medium, wherein instructions in the computer-readable storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the content recommendation method of any one of claims 1-9.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210497195.XA CN114637927B (en) | 2022-05-09 | 2022-05-09 | Content recommendation method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210497195.XA CN114637927B (en) | 2022-05-09 | 2022-05-09 | Content recommendation method and device, electronic equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114637927A CN114637927A (en) | 2022-06-17 |
CN114637927B true CN114637927B (en) | 2022-09-09 |
Family
ID=81952840
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210497195.XA Active CN114637927B (en) | 2022-05-09 | 2022-05-09 | Content recommendation method and device, electronic equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114637927B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116911957B (en) * | 2023-09-12 | 2024-01-05 | 深圳须弥云图空间科技有限公司 | Heterogeneous content mixed recommendation method, device, system and storage medium |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020093289A1 (en) * | 2018-11-07 | 2020-05-14 | 深圳市欢太科技有限公司 | Resource recommendation method and apparatus, electronic device and storage medium |
CN111428133A (en) * | 2020-03-19 | 2020-07-17 | 腾讯科技(北京)有限公司 | Artificial intelligence based recommendation method and device, electronic equipment and storage medium |
CN113886732A (en) * | 2020-07-02 | 2022-01-04 | 腾讯数码(天津)有限公司 | Information recommendation method |
CN114385916A (en) * | 2022-01-11 | 2022-04-22 | 北京达佳互联信息技术有限公司 | Method and device for determining recommended resources, electronic equipment and storage medium |
-
2022
- 2022-05-09 CN CN202210497195.XA patent/CN114637927B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN114637927A (en) | 2022-06-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Yan et al. | AppJoy: personalized mobile application discovery | |
EP3098770A1 (en) | Advertising method and device using cohort-based user analysis platform and marketing platform | |
US11430018B2 (en) | Line item-based audience extension | |
US20190130440A1 (en) | Method and system for detecting fraudulent advertisement activity | |
CN109471978B (en) | Electronic resource recommendation method and device | |
CN107077498B (en) | Representing entity relationships in online advertisements | |
CN105608125B (en) | Information processing method and server | |
CN110210882A (en) | Promote position matching process and device, promotion message methods of exhibiting and device | |
CN113902473B (en) | Training method and device of business prediction system | |
CN109034867B (en) | Click traffic detection method and device and storage medium | |
CN111026969B (en) | Content recommendation method and device, storage medium and server | |
CN113101655A (en) | Virtual prop recommendation method, device, equipment and storage medium | |
CN108512883A (en) | A kind of information-pushing method, device and readable medium | |
CN113536104A (en) | Information recommendation method, device, equipment and storage medium | |
CN114637927B (en) | Content recommendation method and device, electronic equipment and storage medium | |
WO2023116219A1 (en) | Cdn node distribution method and apparatus, electronic device, medium and program product | |
CN111626767B (en) | Resource data issuing method, device and equipment | |
CN111159553A (en) | Information pushing method and device, computer equipment and storage medium | |
WO2022199473A1 (en) | Service analysis method and apparatus based on differential privacy | |
CN111798281B (en) | Method and device for judging cheating operation, computer readable storage medium and equipment | |
CN115375339A (en) | Multimedia information recommendation method, device and equipment and computer storage medium | |
CN113688305A (en) | Information processing method and device and computer readable storage medium | |
CN111193598B (en) | Group chat session recommendation method and device | |
CN111915315B (en) | Authentication mode recommendation method and device, electronic equipment and readable storage medium | |
WO2023027627A1 (en) | Method, device and computer readable medium for rendering organic and paid content |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |