Content recommendation method, apparatus, medium, and electronic device
By acquiring matching features of user attributes and historical clicked content, and utilizing factorization machines and deep neural network models, the accuracy and scalability issues of content recommendation platforms were solved, achieving more accurate content recommendations and more comprehensive feature expression.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING YOUZHUJU NETWORK TECH CO LTD
- Filing Date
- 2023-01-04
- Publication Date
- 2026-06-26
AI Technical Summary
Existing content recommendation platforms struggle to accurately recommend content that users like, and their scalability is poor when handling new content attributes, resulting in insufficient recommendation accuracy and interpretability.
By acquiring the target user's user attribute information and the first content attribute information of the historical clicked content, matching features are determined, and a factorization machine model is used to extract related features. Combined with a deep neural network model, it is determined whether to push the recommended content.
It improves the accuracy and scalability of content recommendations, enhances the flexibility and interpretability of feature combinations, and can better adapt to updates in content attributes.
Smart Images
Figure CN115964566B_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of artificial intelligence technology, and more specifically, to a content recommendation method, apparatus, medium, and electronic device. Background Technology
[0002] With the rapid development of mobile internet and the explosive growth of short video and image consumption, content filtering and recommendation in information feeds are playing an increasingly important role. Content recommendation platforms typically need to recommend hundreds of billions of pieces of content daily, and how to more accurately recommend content that users like is a core metric that these platforms need to continuously optimize. Summary of the Invention
[0003] This summary section is provided to briefly introduce the concepts, which will be described in detail in the detailed description section below. This summary section is not intended to identify key or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.
[0004] In a first aspect, this disclosure provides a content recommendation method, comprising: acquiring user attribute information and historical click content of a target user, and first content attribute information of content to be recommended; determining a matching feature between the target user and the content to be recommended based on the first content attribute information and the historical click content, wherein the matching feature is used to characterize the matching situation of the content attributes of the historical click content and the content to be recommended; extracting at least a first association feature between the first content attribute information and the matching feature; determining whether to push the content to be recommended based on the first association feature, the user attribute information, and the first content attribute information; and if it is determined to push the content to be recommended, then pushing the content to be recommended to the target user.
[0005] Optionally, before the step of extracting at least the first correlation feature between the first content attribute information and the matching feature, the method further includes: determining second content attribute information for each of the historical clicked contents; merging each of the second content attribute information according to different attributes to obtain the behavioral features of the target user; the step of extracting at least the first correlation feature between the first content attribute information and the matching feature includes: extracting the first correlation feature between the first content attribute information, the matching feature, and the behavioral feature.
[0006] Optionally, determining the matching features between the target user and the content to be recommended based on the first content attribute information and the historical clicked content includes: determining the second content attribute information for each of the historical clicked content; for each attribute value in the first content attribute information, determining the number of historical clicked content in the second content attribute information that contains that attribute value; and determining each of the aforementioned numbers as the matching features between the target user and the content to be recommended.
[0007] Optionally, determining whether to push the content to be recommended based on the first association feature, the user attribute information, and the first content attribute information includes: determining the matching degree between the content to be recommended and the target user based on the first association feature, the user attribute information, and the first content attribute information; and determining whether to push the content to be recommended based on the matching degree.
[0008] Optionally, determining the matching degree between the content to be recommended and the target user based on the first association feature, the user attribute information, and the first content attribute information includes: inputting the first association feature, the user attribute information, and the first content attribute information into a deep neural network model to obtain the matching degree between the content to be recommended and the target user.
[0009] Optionally, determining the matching degree between the content to be recommended and the target user based on the first association feature, the user attribute information, and the first content attribute information includes: extracting a second association feature between the user attribute information and the first content attribute information; and inputting the first association feature, the second association feature, the user attribute information, and the first content attribute information into a deep neural network model to obtain the matching degree between the content to be recommended and the target user.
[0010] Optionally, the step of extracting at least the first association feature between the first content attribute information and the matching feature includes: using a first factorization machine model to extract at least the first association feature between the first content attribute information and the matching feature.
[0011] Secondly, this disclosure provides a content recommendation device, comprising: an acquisition module, configured to acquire user attribute information and historical click content of a target user, and first content attribute information of content to be recommended; a first determination module, configured to determine a matching feature between the target user and the content to be recommended based on the first content attribute information and the historical click content, wherein the matching feature is used to characterize the content attribute matching situation between the historical click content and the content to be recommended; an extraction module, configured to extract at least a first association feature between the first content attribute information and the matching feature; a second determination module, configured to determine whether to push the content to be recommended based on the first association feature, the user attribute information, and the first content attribute information; and a push module, configured to push the content to be recommended to the target user if it is determined that the content to be recommended should be pushed.
[0012] Thirdly, this disclosure provides a computer-readable medium having a computer program stored thereon, which, when executed by a processing device, implements the steps of the content recommendation method provided in the first aspect of this disclosure.
[0013] Fourthly, this disclosure provides an electronic device, comprising: a storage device having a computer program stored thereon; and a processing device for executing the computer program in the storage device to implement the steps of the content recommendation method provided in the first aspect of this disclosure.
[0014] In the above technical solution, firstly, based on the first content attribute information of the content to be recommended and the target user's historical click content, the matching features between the target user and the content to be recommended are determined. Then, at least the first association feature between the first content attribute information and the matching features is extracted. Next, based on the first association feature, user attribute information, and first content attribute information, it is determined whether to push the content to be recommended. If it is determined to push the content to be recommended, then the content to be recommended is pushed to the target user. By extracting the association features between the content attributes and matching features of the content to be recommended, the feature combination logic between users and content can be enhanced, the feature combination expression is more complete and detailed, and the flexibility of the combination interaction of underlying features is improved, thereby enabling more accurate push of content that users like. In addition, even if the content attributes of the content to be recommended are updated, the new content attributes can still have the ability to combine features, making the feature combination expression more complete, thereby improving the scalability of content attributes. Furthermore, determining the association features based on content attributes enhances the interpretability of feature expression, allowing the importance of various features to be reflected in the importance of specific content attribute features, thereby enhancing the interpretability of the content recommendation platform.
[0015] Other features and advantages of this disclosure will be described in detail in the following detailed description section. Attached Figure Description
[0016] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale. In the drawings:
[0017] Figure 1 This is a flowchart illustrating a content recommendation method according to an exemplary embodiment.
[0018] Figure 2 This is a flowchart illustrating a content recommendation method according to another exemplary embodiment.
[0019] Figure 3 This is a schematic diagram illustrating a content recommendation method according to an exemplary embodiment.
[0020] Figure 4 This is a block diagram illustrating a content recommendation apparatus according to an exemplary embodiment.
[0021] Figure 5 This is a schematic diagram of the structure of an electronic device according to an exemplary embodiment. Detailed Implementation
[0022] As discussed in the background section, how to more accurately recommend content that users like is a core metric that content recommendation platforms need to continuously optimize. Currently, content recommendation platforms typically use Factorization Machines (FM) to perform interactive operations on the content attribute features of the content to be recommended and the user attribute features to obtain their correlation features. Then, based on the content attribute features, user attribute features, and correlation features, a deep neural network model is used to obtain the matching degree between the user and the content to be recommended, and the matching degree is used to determine whether to push corresponding content to the user. Specifically, when performing interactive operations on the content attribute features and user attribute features, feature groups to be calculated for correlation are first manually enumerated; then, the correlation features of each feature group are calculated using FM. However, since the enumerated feature groups are limited, the flexibility of the underlying feature combination interaction is reduced. Furthermore, since the content attribute features of the content to be recommended are usually few, a single content attribute feature may be reused multiple times in different feature groups, failing to fully explore the feature information of a single content attribute feature. This will affect the accuracy of the matching degree calculation and make it difficult to guarantee the accuracy of the recommended content. In addition, manually enumerating feature groups is not conducive to adding new content attributes, resulting in poor scalability.
[0023] In view of this, the present disclosure provides a content recommendation method, apparatus, medium and electronic device.
[0024] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.
[0025] It should be understood that the steps described in the method embodiments of this disclosure may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of this disclosure is not limited in this respect.
[0026] The term "comprising" and its variations as used herein are open-ended inclusions, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.
[0027] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.
[0028] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".
[0029] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.
[0030] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.
[0031] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.
[0032] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.
[0033] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.
[0034] Meanwhile, it is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.
[0035] Figure 1 This is a flowchart illustrating a content recommendation method according to an exemplary embodiment. For example... Figure 1 As shown, the method may include the following S101 to S104.
[0036] In S101, the user attribute information and historical click content of the target user are obtained, as well as the first content attribute information of the content to be recommended.
[0037] In this disclosure, the target user is any user to whom content recommendations are to be made, and the user attribute information is the user's basic attribute information. Historical click content may include the user's content click records within a preset historical period (e.g., the most recent month).
[0038] The content to be recommended can be content that the content recommendation platform wants to recommend to target users. The first content attribute information of the content to be recommended can include information that can characterize the theme of the content to be recommended. The first content attribute information can include, but is not limited to, content ID, keywords, category tags (e.g., news, pets, sports, etc.), and content type (e.g., video, audio, article, etc.). Each content attribute in the first content attribute information may have different dimensions. For example, keywords can include title keywords, voice keywords, and Optical Character Recognition (OCR) keywords.
[0039] The aforementioned content recommendation platforms can be media content recommendation platforms, including but not limited to news recommendation platforms, music recommendation platforms, and video recommendation platforms.
[0040] In S102, the matching characteristics between the target user and the content to be recommended are determined based on the first content attribute information and the historical click content.
[0041] In this disclosure, the matching features between the target user and the content to be recommended are used to characterize the matching of the target user's historical click content with the content attributes of the content to be recommended.
[0042] In S103, at least the first association feature between the first content attribute information and the matching feature is extracted.
[0043] In S104, based on the first association feature, user attribute information, and first content attribute information, it is determined whether to push the content to be recommended.
[0044] In S105, if it is determined that content to be recommended will be pushed, then the content to be recommended will be pushed to the target user.
[0045] In the above technical solution, firstly, based on the first content attribute information of the content to be recommended and the target user's historical click content, the matching features between the target user and the content to be recommended are determined. Then, at least the first association feature between the first content attribute information and the matching features is extracted. Next, based on the first association feature, user attribute information, and first content attribute information, it is determined whether to push the content to be recommended. If it is determined to push the content to be recommended, then the content to be recommended is pushed to the target user. By extracting the association features between the content attributes and matching features of the content to be recommended, the feature combination logic between users and content can be enhanced, the feature combination expression is more complete and detailed, and the flexibility of the combination interaction of underlying features is improved, thereby enabling more accurate push of content that users like. In addition, even if the content attributes of the content to be recommended are updated, the new content attributes can still have the ability to combine features, making the feature combination expression more complete, thereby improving the scalability of content attributes. Furthermore, determining the association features based on content attributes enhances the interpretability of feature expression, allowing the importance of various features to be reflected in the importance of specific content attribute features, thereby enhancing the interpretability of the content recommendation platform.
[0046] The following is a detailed description of the specific implementation method for determining the matching features between the target user and the content to be recommended based on the first content attribute information and historical click content in S102 above. Specifically, it can be achieved through the following steps [1] to [3]:
[0047] Step [1]: Determine the second content attribute information for each historical click content.
[0048] The second content attribute information of the historical clicked content may include information that can characterize the theme of the historical clicked content. The second content attribute information may include, but is not limited to, content ID, keywords, category tags, and content subject type. Each content attribute may have different dimensions.
[0049] Step [2]: For each attribute value in the first content attribute information, determine the number of historical clicked contents containing that attribute value in the second content attribute information.
[0050] For example, if the target user has 100 historical clicks, and the "tag" content attribute in the first content attribute information has the attribute value "pets", then 11 of the 100 historical clicks have the tag "pets". Therefore, the number of historical clicks with the attribute value "pets" in the second content attribute information is 11.
[0051] Step [3]: Determine each quantity as a matching feature between the target user and the content to be recommended.
[0052] The following is a detailed description of the specific implementation of at least extracting the first correlation feature between the first content attribute information and the matching feature in S103 above.
[0053] In one implementation, a first association feature is extracted between the first content attribute information and the matching feature.
[0054] Specifically, a first factorization machine model can be used to extract the first association feature between the first content attribute information and the matching feature. In this case, the first association feature can be the cross-dot product feature of the first content attribute information and the matching feature.
[0055] For example, for each content attribute in the first content attribute information, the association feature between the content attribute and the corresponding matching feature can be calculated using the following equation (1):
[0056]
[0057] in, Attr is the associated feature (feature vector) between the k-th content attribute in the first content attribute information and the corresponding matching feature; i Let be the feature vector of the i-th attribute value of the k-th content attribute, where a content attribute can have attribute values with different dimensions, i = 1, 2, ..., I, where I is the number of dimensions of the k-th content attribute, I ≥ 1; AttrMatch m W is the feature vector corresponding to the number of historical clicked content items containing the m-th attribute value of the k-th content attribute in the second content attribute information. m = 1, 2, ..., I; Ni For Attr i The corresponding weight, W Mm For AttrMatch m The corresponding weights are all model parameters of the first factorization machine model.
[0058] In another implementation, such as Figure 2 As shown, before S103 above, the above method may also include the following S106 and S107.
[0059] In S106, the second content attribute information of each historical click content is determined.
[0060] In S107, each second content attribute information is merged according to different attributes to obtain the behavioral characteristics of the target user.
[0061] In this disclosure, merging the various second content attribute information according to different attributes means calculating the union of the attribute values of each content attribute in the various second content attribute information for each content attribute in the second content attribute information.
[0062] At this point, S103 may include: extracting a first association feature among the first content attribute information, matching features, and behavioral features. Specifically, a first factorization machine model can be used to extract the first association feature among the first content attribute information, matching features, and behavioral features. In this case, the first association feature can be the cross-dot product feature of the first content attribute information, matching features, and behavioral features.
[0063] For example, for each content attribute of the first content attribute information, the association features of the content attribute, the matching features corresponding to the content attribute, and the behavioral features corresponding to the content attribute can be calculated using the following equation (2):
[0064]
[0065] in, The associated features (as feature vectors) are the k-th content attribute, the matching feature corresponding to the k-th content attribute, and the behavioral feature corresponding to the k-th content attribute in the first content attribute information; AttrList j Let W be the feature vector corresponding to the j-th attribute value in the behavioral features corresponding to the k-th content attribute. The behavioral feature corresponding to the k-th content attribute is the attribute value obtained by merging the second content attribute information of each historical clicked content according to the k-th content attribute, j = 1, 2, ..., J, where J is the number of attribute values obtained by merging the second content attribute information of each historical clicked content according to the k-th content attribute, J ≥ 1; Lj For AttrList j The corresponding weights are the model parameters of the first factorization machine model.
[0066] The following is a detailed description of the specific implementation method for determining whether to push recommended content based on the first association feature, user attribute information, and first content attribute information in S104 above. Specifically, it can be achieved through the following steps (1) and (2):
[0067] Step (1): Determine the matching degree between the content to be recommended and the target user based on the first association feature, user attribute information and the first content attribute information.
[0068] In one implementation, the first association feature, user attribute information, and first content attribute information can be input into a deep neural network model to obtain the matching degree between the content to be recommended and the target user.
[0069] In another implementation, such as Figure 3 As shown, the second correlation feature between the user attribute information and the first content attribute information of the target user can be extracted first. Then, the first correlation feature, the second correlation feature, the user attribute information and the first content attribute information are input into the deep neural network model to obtain the matching degree between the content to be recommended and the target user.
[0070] Step (2): Determine whether to push the recommended content based on the matching degree between the content to be recommended and the target user.
[0071] Specifically, if the matching degree between the content to be recommended and the target user is greater than or equal to the preset matching degree threshold, then the content to be recommended will be pushed; if the matching degree between the content to be recommended and the target user is less than the preset matching degree threshold, then the content to be recommended will not be pushed.
[0072] The following provides a detailed description of the specific implementation method for extracting the second correlation feature between the user attribute information and the first content attribute information of the target user. Specifically, a second factorization machine model can be used to extract the second correlation feature between the user attribute information and the first content attribute information of the target user (e.g., ...). Figure 3 As shown in the figure, the second association feature can be the cross-dot product feature of the target user's user attribute information and the first content attribute information. Before extracting the second association feature, the second factorization machine model first manually enumerates the feature groups to be calculated, that is, selecting at least one user attribute from the target user's user attribute information and at least one content attribute from the first content attribute information to form a feature group; then, it calculates the association feature for each feature group.
[0073] For example, for each feature group in the manually enumerated feature group to be used for the second association feature calculation, the association feature of that feature group can be calculated using the following equation (3):
[0074] FM_r=(∑p User p *W Up )*(∑ q Item q *W Xq (3)
[0075] Wherein, FM_r represents the associated feature of the r-th feature group in the feature group to be calculated for the second associated feature, which is manually enumerated; User p Let Item be the p-th user attribute in the r-th feature group, where p = 1, 2, ..., P, and P is the total number of user attributes in the r-th feature group, P ≥ 1; q Let W be the q-th content attribute in the r-th feature group, where q = 1, 2, ..., Q, and Q is the total number of content attributes in the r-th feature group, Q ≥ 1; Up For User p The corresponding weight, W Xq For Item q The corresponding weights are all model parameters of the second factorization machine model.
[0076] Figure 4 This is a block diagram illustrating a content recommendation apparatus according to an exemplary embodiment. Figure 4 As shown, the device 400 includes: an acquisition module 401, configured to acquire user attribute information and historical click content of a target user, and first content attribute information of content to be recommended; a first determination module 402, configured to determine the matching features between the target user and the content to be recommended based on the first content attribute information and the historical click content, wherein the matching features are used to characterize the content attribute matching situation between the historical click content and the content to be recommended; an extraction module 403, configured to extract at least a first association feature between the first content attribute information and the matching features; a second determination module 404, configured to determine whether to push the content to be recommended based on the first association feature, the user attribute information, and the first content attribute information; and a push module 405, configured to push the content to be recommended to the target user if it is determined that the content to be recommended should be pushed.
[0077] In the above technical solution, firstly, based on the first content attribute information of the content to be recommended and the target user's historical click content, the matching features between the target user and the content to be recommended are determined. Then, at least the first association feature between the first content attribute information and the matching features is extracted. Next, based on the first association feature, user attribute information, and first content attribute information, it is determined whether to push the content to be recommended. If it is determined to push the content to be recommended, then the content to be recommended is pushed to the target user. By extracting the association features between the content attributes and matching features of the content to be recommended, the feature combination logic between users and content can be enhanced, the feature combination expression is more complete and detailed, and the flexibility of the combination interaction of underlying features is improved, thereby enabling more accurate push of content that users like. In addition, even if the content attributes of the content to be recommended are updated, the new content attributes can still have the ability to combine features, making the feature combination expression more complete, thereby improving the scalability of content attributes. Furthermore, determining the association features based on content attributes enhances the interpretability of feature expression, allowing the importance of various features to be reflected in the importance of specific content attribute features, thereby enhancing the interpretability of the content recommendation platform.
[0078] Optionally, the device 400 further includes: a third determining module, configured to determine second content attribute information for each of the historical clicked contents before the extraction module 403 extracts at least the first correlation feature between the first content attribute information and the matching feature; a merging module, configured to merge each of the second content attribute information according to different attributes to obtain the behavioral features of the target user; and the extraction module 403 is configured to extract the first content attribute information, the matching feature, and the first correlation feature between the behavioral features.
[0079] Optionally, the first determining module 402 includes: a first determining submodule, configured to determine second content attribute information for each of the historical clicked contents; a second determining submodule, configured to determine, for each attribute value in the first content attribute information, the number of historical clicked contents in the second content attribute information that contain that attribute value; and a third determining submodule, configured to determine each of the quantities as a matching feature between the target user and the content to be recommended.
[0080] Optionally, the second determining module 404 includes: a fourth determining submodule, configured to determine the matching degree between the content to be recommended and the target user based on the first association feature, the user attribute information, and the first content attribute information; and a fifth determining submodule, configured to determine whether to push the content to be recommended based on the matching degree.
[0081] Optionally, the fourth determining submodule is used to input the first association feature, the user attribute information, and the first content attribute information into a deep neural network model to obtain the matching degree between the content to be recommended and the target user.
[0082] Optionally, the fourth determining submodule includes: an extraction submodule, used to extract a second association feature between the user attribute information and the first content attribute information; and a matching degree calculation submodule, used to input the first association feature, the second association feature, the user attribute information, and the first content attribute information into a deep neural network model to obtain the matching degree between the content to be recommended and the target user.
[0083] Optionally, the extraction module 403 is used to extract at least a first association feature between the first content attribute information and the matching feature using a first factorization machine model.
[0084] This disclosure also provides a computer-readable medium having a computer program stored thereon, which, when executed by a processing device, implements the steps of the method recommended above provided in this disclosure.
[0085] The following is for reference. Figure 5 The diagram illustrates a structural schematic of an electronic device (e.g., a terminal device or a server) 600 suitable for implementing embodiments of the present disclosure. The terminal device in the embodiments of the present disclosure may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 5 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.
[0086] like Figure 5 As shown, electronic device 600 may include a processing device (e.g., a central processing unit, a graphics processor, etc.) 601, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 602 or a program loaded from storage device 608 into random access memory (RAM) 603. RAM 603 also stores various programs and data required for the operation of electronic device 600. Processing device 601, ROM 602, and RAM 603 are interconnected via bus 604. Input / output (I / O) interface 605 is also connected to bus 604.
[0087] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 608 including, for example, magnetic tapes, hard disks, etc.; and communication devices 609. Communication device 609 allows electronic device 600 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 5 An electronic device 600 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively.
[0088] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a storage device 608, or installed from a ROM 602. When the computer program is executed by the processing device 601, it performs the functions defined in the methods of embodiments of this disclosure.
[0089] It should be noted that the computer-readable medium described in this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.
[0090] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.
[0091] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device.
[0092] The aforementioned computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to: acquire user attribute information and historical click content of a target user, and first content attribute information of content to be recommended; determine matching features between the target user and the content to be recommended based on the first content attribute information and the historical click content, wherein the matching features characterize the content attribute matching between the historical click content and the content to be recommended; extract at least a first association feature between the first content attribute information and the matching features; determine whether to push the content to be recommended based on the first association feature, the user attribute information, and the first content attribute information; and if it is determined that the content to be recommended should be pushed, then push the content to be recommended to the target user.
[0093] Computer program code for performing the operations of this disclosure can be written in one or more programming languages or a combination thereof, including but not limited to object-oriented programming languages such as Java, Smalltalk, and C++, as well as conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0094] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.
[0095] The modules described in the embodiments of this disclosure can be implemented in software or hardware. The names of the modules do not necessarily limit the module itself; for example, the first determining module can also be described as "a module that determines the matching features between the target user and the content to be recommended based on the first content attribute information and the historical click content".
[0096] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.
[0097] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.
[0098] According to one or more embodiments of this disclosure, Example 1 provides a content recommendation method, including: obtaining user attribute information and historical click content of a target user, and first content attribute information of content to be recommended; determining a matching feature between the target user and the content to be recommended based on the first content attribute information and the historical click content, wherein the matching feature is used to characterize the matching situation of the content attributes of the historical click content and the content to be recommended; extracting at least a first association feature between the first content attribute information and the matching feature; determining whether to push the content to be recommended based on the first association feature, the user attribute information, and the first content attribute information; if it is determined to push the content to be recommended, then pushing the content to be recommended to the target user.
[0099] According to one or more embodiments of this disclosure, Example 2 provides the method of Example 1, which, before the step of at least extracting the first association feature between the first content attribute information and the matching feature, further includes: determining second content attribute information for each of the historical clicked contents; merging each of the second content attribute information according to different attributes to obtain the behavioral feature of the target user; the step of at least extracting the first association feature between the first content attribute information and the matching feature includes: extracting the first association feature between the first content attribute information, the matching feature, and the behavioral feature.
[0100] According to one or more embodiments of this disclosure, Example 3 provides the method of Example 1, wherein determining the matching features between the target user and the content to be recommended based on the first content attribute information and the historical clicked content includes: determining second content attribute information for each of the historical clicked content; for each attribute value in the first content attribute information, determining the number of historical clicked content in the second content attribute information that contains that attribute value; and determining each of the quantities as the matching feature between the target user and the content to be recommended.
[0101] According to one or more embodiments of this disclosure, Example 4 provides a method according to any one of Examples 1-3, wherein determining whether to push the content to be recommended based on the first association feature, the user attribute information, and the first content attribute information includes: determining the matching degree between the content to be recommended and the target user based on the first association feature, the user attribute information, and the first content attribute information; and determining whether to push the content to be recommended based on the matching degree.
[0102] According to one or more embodiments of this disclosure, Example 5 provides the method of Example 4, wherein determining the matching degree between the content to be recommended and the target user based on the first association feature, the user attribute information and the first content attribute information includes: inputting the first association feature, the user attribute information and the first content attribute information into a deep neural network model to obtain the matching degree between the content to be recommended and the target user.
[0103] According to one or more embodiments of this disclosure, Example 6 provides the method of Example 4, wherein determining the matching degree between the content to be recommended and the target user based on the first association feature, the user attribute information, and the first content attribute information includes: extracting a second association feature between the user attribute information and the first content attribute information; and inputting the first association feature, the second association feature, the user attribute information, and the first content attribute information into a deep neural network model to obtain the matching degree between the content to be recommended and the target user.
[0104] According to one or more embodiments of this disclosure, Example 7 provides a method according to any one of Examples 1-3, wherein the step of extracting at least a first association feature between the first content attribute information and the matching feature includes: using a first factorization machine model to extract at least a first association feature between the first content attribute information and the matching feature.
[0105] According to one or more embodiments of this disclosure, Example 8 provides a content recommendation apparatus, comprising: an acquisition module, configured to acquire user attribute information and historical click content of a target user, and first content attribute information of content to be recommended; a first determination module, configured to determine a matching feature between the target user and the content to be recommended based on the first content attribute information and the historical click content, wherein the matching feature is used to characterize the matching situation of the content attributes of the historical click content and the content to be recommended; an extraction module, configured to extract at least a first association feature between the first content attribute information and the matching feature; a second determination module, configured to determine whether to push the content to be recommended based on the first association feature, the user attribute information, and the first content attribute information; and a push module, configured to push the content to be recommended to the target user if it is determined that the content to be recommended should be pushed.
[0106] According to one or more embodiments of the present disclosure, Example 9 provides a computer-readable medium having a computer program stored thereon that, when executed by a processing device, implements the steps of the method described in any one of Examples 1-7.
[0107] According to one or more embodiments of this disclosure, Example 10 provides an electronic device including: a storage device having a computer program stored thereon; and a processing device for executing the computer program in the storage device to implement the steps of any one of Examples 1-7.
[0108] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features disclosed in this disclosure that have similar functions.
[0109] Furthermore, while the operations are described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or in a sequential order. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.
[0110] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative forms of implementing the claims. Regarding the apparatus in the above embodiments, the specific manner in which the various modules perform their operations has been described in detail in the embodiments relating to the method, and will not be elaborated upon here.
Claims
1. A content recommendation method, characterized in that, include: Obtain the target user's user attribute information and historical click content, as well as the first content attribute information of the content to be recommended; Based on the first content attribute information and the historical click content, the matching features between the target user and the content to be recommended are determined, wherein the matching features are used to characterize the matching of the content attributes between the historical click content and the content to be recommended; At least a first association feature between the first content attribute information and the matching feature is extracted, wherein the first association feature is the cross-dot product feature of the first content attribute information and the matching feature; Based on the first association feature, the user attribute information, and the first content attribute information, determine whether to push the content to be recommended; If it is determined that the content to be recommended will be pushed, then the content to be recommended will be pushed to the target user; The step of determining the matching features between the target user and the content to be recommended based on the first content attribute information and the historical click content includes: Determine the second content attribute information for each of the historical clicked contents; For each attribute value in the first content attribute information, determine the number of historical clicked contents containing that attribute value in the second content attribute information; Each of the quantities is determined as a matching feature between the target user and the content to be recommended.
2. The method according to claim 1, characterized in that, Prior to the step of extracting at least the first association feature between the first content attribute information and the matching feature, the method further includes: Each of the second content attribute information is merged according to different attributes to obtain the behavioral characteristics of the target user; The step of extracting at least the first association feature between the first content attribute information and the matching feature includes: Extract the first association feature among the first content attribute information, the matching feature, and the behavioral feature.
3. The method according to any one of claims 1-2, characterized in that, The step of determining whether to push the content to be recommended based on the first association feature, the user attribute information, and the first content attribute information includes: Based on the first association feature, the user attribute information, and the first content attribute information, the matching degree between the content to be recommended and the target user is determined; Based on the matching degree, determine whether to push the recommended content.
4. The method according to claim 3, characterized in that, The step of determining the matching degree between the content to be recommended and the target user based on the first association feature, the user attribute information, and the first content attribute information includes: The first association feature, the user attribute information, and the first content attribute information are input into a deep neural network model to obtain the matching degree between the content to be recommended and the target user.
5. The method according to claim 3, characterized in that, The step of determining the matching degree between the content to be recommended and the target user based on the first association feature, the user attribute information, and the first content attribute information includes: Extract the second association feature between the user attribute information and the first content attribute information; The first association feature, the second association feature, the user attribute information, and the first content attribute information are input into a deep neural network model to obtain the matching degree between the content to be recommended and the target user.
6. The method according to any one of claims 1-2, characterized in that, The step of extracting at least the first association feature between the first content attribute information and the matching feature includes: The first factorization machine model is used to extract at least the first association feature between the first content attribute information and the matching feature.
7. A content recommendation device, characterized in that, include: The acquisition module is used to acquire the user attribute information and historical click content of the target user, as well as the first content attribute information of the content to be recommended; The first determining module is used to determine the matching features between the target user and the content to be recommended based on the first content attribute information and the historical click content, wherein the matching features are used to characterize the matching situation of the content attributes of the historical click content and the content to be recommended; An extraction module is configured to extract at least a first association feature between the first content attribute information and the matching feature, wherein the first association feature is the cross-dot product feature of the first content attribute information and the matching feature; The second determining module is used to determine whether to push the content to be recommended based on the first association feature, the user attribute information, and the first content attribute information. The push module is used to push the recommended content to the target user if it is determined that the recommended content will be pushed. The first determining module includes: The first determining submodule is used to determine the second content attribute information of each of the historical clicked contents; The second determining submodule is used to determine the number of historical clicked contents in the second content attribute information that contain each attribute value for each attribute value in the first content attribute information. The third determining submodule is used to determine each of the quantities as a matching feature between the target user and the content to be recommended.
8. A computer-readable medium having a computer program stored thereon, characterized in that, When executed by the processing device, the program implements the steps of the method according to any one of claims 1-6.
9. An electronic device, characterized in that, include: A storage device on which computer programs are stored; A processing device for executing the computer program in the storage device to implement the steps of the method according to any one of claims 1-6.