Method and device for determining to-be-recommended resources based on large model, equipment and storage medium
By obtaining user behavior preferences and refresh depth, combining historical resource information, and using large models to filter candidate resources, the problem of inaccurate recommendations in existing recommendation systems at different refresh depths is solved, and the user experience is improved.
Patent Information
- Application Number
- CN202411311794.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-19
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-09-19
AI Technical Summary
In continuous playback scenarios, it is difficult for the existing recommendation system to accurately match the user's interest changes under different refresh depths, resulting in insufficient resource recommendations and affecting the user experience.
By obtaining the current user's behavioral preferences and refresh depth, combining the playback attribute information of historical resources, using a large model to determine resource satisfaction information, and filtering candidate resources based on this to improve the matching between the resources to be recommended and the user's needs.
It improves the matching degree between the resources to be recommended and the user needs, and improves the user experience, especially when the user refreshes the depth of the recommended content.
Smart Images

Figure CN119089049A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the computer field, in particular to the field of artificial intelligence technology, deep learning, multimedia processing and other technical fields, and specifically to a method, device, equipment and storage medium for determining resources to be recommended based on a large model. Background Art
[0002] With the rapid advancement of information technology and the widespread application of big data and artificial intelligence technologies, recommendation systems are booming at an unprecedented rate, especially in short video applications, showing great potential and vitality. Users only need to simply enter their behavioral preferences or conduct preliminary exploration based on their personal interests and preferences, and the recommendation system can immediately analyze massive data and accurately push short video content that suits their tastes. While enjoying the pleasant experience brought by personalized recommendations, users can also refresh the page continuously to seamlessly connect to the next exciting video, achieving a smooth experience of continuous playback. How to recommend resources in a continuous playback scenario is very important. Summary of the invention
[0003] The present disclosure provides a method, apparatus, device and storage medium for determining resources to be recommended based on a large model.
[0004] According to one aspect of the present disclosure, a method for determining resources to be recommended based on a large model is provided, comprising:
[0005] Get the current user's behavior preferences and current refresh depth;
[0006] Determining, based on the behavior preference and the initial resource, a similarity between the candidate resource and the candidate resource;
[0007] Determine the resource satisfaction information of the current user at the current refresh depth according to the current refresh depth, the historical resources played by the current user and the playback attribute information of the historical resources; wherein the playback attribute information at least includes the historical refresh depth;
[0008] The candidate resources are screened according to the resource satisfaction information and the similarity, and the recommended resources for the current user at the current refresh depth are determined based on the screened candidate resources based on the big model.
[0009] According to one aspect of the present disclosure, there is provided a device for determining resources to be recommended based on a large model, comprising:
[0010] The current acquisition module is used to obtain the current user's behavior preferences and current refresh depth;
[0011] A candidate resource module, used to determine the similarity between a candidate resource and the candidate resource according to the behavior preference and the initial resource;
[0012] A resource satisfaction module, used to determine the resource satisfaction information of the current user at the current refresh depth according to the current refresh depth, the historical resources played by the current user and the playback attribute information of the historical resources; wherein the playback attribute information at least includes the historical refresh depth;
[0013] The resource module to be recommended is used to screen the candidate resources according to the resource satisfaction information and the similarity, and determine the resource to be recommended for the current user at the current refresh depth based on the screened candidate resources based on the big model.
[0014] According to another aspect of the present disclosure, there is provided an electronic device, the electronic device comprising:
[0015] at least one processor; and
[0016] a memory communicatively connected to the at least one processor; wherein,
[0017] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method provided by any embodiment of the present disclosure.
[0018] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable a computer to execute the method provided by any embodiment of the present disclosure.
[0019] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, wherein when the computer program is executed by a processor, the method provided according to any embodiment of the present disclosure is implemented.
[0020] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] Figure 1 is an architecture diagram of a remote log system provided according to an embodiment of the present disclosure;
[0022] Figure 2 is a flowchart of another method for determining resources to be recommended based on a large model according to an embodiment of the present disclosure;
[0023] Figure 3 is a flowchart of another method for determining resources to be recommended based on a large model according to an embodiment of the present disclosure;
[0024] Figure 4It is a structural schematic diagram of a device for determining resources to be recommended based on a large model according to an embodiment of the present disclosure;
[0025] Figure 5 The block diagram of an electronic device is used to implement the method for determining resources to be recommended based on a large model according to an embodiment of the present disclosure. DETAILED DESCRIPTION
[0026] Figure 1 The flowchart of a method for determining resources to be recommended based on a large model according to an embodiment of the present disclosure is provided. The method is applicable to the case of pushing resources to users. The method can be executed by a device for determining resources to be recommended based on a large model, and the device can be implemented in software and / or hardware and can be integrated into an electronic device. Figure 1 As shown, the method for determining resources to be recommended based on a large model in this embodiment may include:
[0027] S101, obtaining the current user's behavior preference and current refresh depth;
[0028] S102, determining the similarity between a candidate resource and the candidate resource according to the behavior preference and the initial resource;
[0029] S103, determining resource satisfaction information of the current user at the current refresh depth according to the current refresh depth, the historical resources played by the current user, and the playback attribute information of the historical resources; wherein the playback attribute information at least includes the historical refresh depth;
[0030] S104, screening the candidate resources according to the resource satisfaction information and the similarity, and determining the recommended resources for the current user at the current refresh depth based on the screened candidate resources based on the big model.
[0031] The current user refers to the user who currently needs to make resource recommendations, that is, the user who needs to play the resource. The current user's behavioral preferences are used to characterize the current user's behavioral characteristics, interests, historical choices, or personal preferences. The current refresh depth is used to characterize the depth of continuous playback or sequence continuity of the current user in the current playback session, reflecting the continuity and depth of exploration of the current user during playback. The current playback session can be started when the user starts the resource application or the user enters a query request. If the user enters the query request in text or voice, the query request belongs to the behavioral characteristics of the current user.
[0032] For example, in the current playback session, resources can be pushed to the current user according to the current user's behavior preference, the first resource played in the current playback session is obtained, and the new resource is played seamlessly through the refresh operation. For the resource played by the current user for the i-th time in the current playback session (i is a positive integer), the corresponding current refresh depth is i, where i is a positive integer.
[0033] Historical resources are resources played by the current user in historical playback sessions. Historical refresh depth is used to characterize the depth or sequence continuity of continuous playback by the current user in historical playback sessions. The historical playback session is before the current playback session. Resource satisfaction information is used to characterize the current user's satisfaction with resources at the current refresh depth. It is the resource interest signal of the current user based on the current refresh depth, and is used to more accurately and timely grasp the current user's interest preference for resources at the current refresh depth.
[0034] Exemplarily, the behavior preference of the current user may be encoded to obtain a vector representation (embedding) of the behavior preference of the current user, the semantic similarity between the behavior preference vector representation and the vector representation of each initial resource may be determined, and multiple initial resources with relatively high semantic similarity may be used as candidate resources, and the similarity of the candidate resources may be obtained. The historical resources played by the current user in the historical playback session and the historical refresh depth of the historical resources may be obtained to obtain playback attribute information including at least the historical refresh depth. The playback attribute information of the historical resources may also include the viewing time, completion or interactive behavior of the historical resources.
[0035] Exemplarily, whether the current user is satisfied with the historical resources can be determined based on the playback attribute information of the historical resources to obtain the satisfaction information of the current user with the historical resources; the resource satisfaction information of the current refresh depth can be determined by combining the current refresh depth, the historical resource depth, and the satisfaction information with the historical resources. Moreover, the resource satisfaction information of the current refresh depth and the similarity of each candidate resource can be used to screen each candidate resource, and based on the large model, the recommended resources can be determined for the current user at the current refresh depth from the candidate resources retained after screening. For example, the large language model can be used to sort the candidate resources after screening, and the recommended resources can be determined based on the sorting results for the current user to play at the current refresh depth.
[0036] In the resource recommendation scenario, users will gradually become fatigued as the refresh depth increases. Specifically, as the refresh depth increases, the resource completion rate, effective playback rate, resource playback time, etc. show a gradually decreasing trend, and the fast scrolling rate gradually increases. Therefore, users have different interests at different refresh depths. In the disclosed embodiment, the resource satisfaction information is determined by accurately integrating the current refresh depth, historical resources, and the playback attribute information of the historical resources. The process of determining the resource satisfaction information not only takes into account the current refresh depth in the current user's immediate behavior, but also takes into account the long-term preferences and habits of the current user through the playback attribute information of historical resources, so that the resource satisfaction information of the current user at the current refresh depth can be obtained more accurately. Moreover, by introducing the resource satisfaction information of the current refresh depth as an interest signal into the resource recommendation process, the resources to be recommended are determined for the current user at the current refresh depth according to the resource satisfaction information and the similarity of each candidate resource, which improves the matching degree between the resources to be recommended and the needs of the current user, thereby effectively improving the user experience.
[0037] The technical solution provided by the embodiment of the present disclosure obtains the current refresh depth of the current user in the current playback session, the historical resources in the historical playback session and the playback attribute information of the historical resources, and combines and integrates the current refresh depth, historical resources and the playback attribute information of the historical resources to determine the resource satisfaction information of the current user at the current refresh depth, and introduces the resource satisfaction information of the current refresh depth as an interest signal into the resource recommendation process, thereby improving the matching degree between the resources to be recommended and the current user needs, thereby effectively improving the user experience.
[0038] Figure 2 FIG. 1 is a flowchart of another method for determining resources to be recommended based on a large model according to an embodiment of the present disclosure. Figure 2 Based on the above embodiment, the method of determining resources to be recommended based on a large model in this embodiment may include:
[0039] S201, obtaining the current user's behavior preference and current refresh depth;
[0040] S202, determining the similarity between the candidate resource and the candidate resource according to the behavior preference and the initial resource;
[0041] S203, determining a reference resource associated with the current refresh depth from the historical resources according to the current refresh depth and the historical refresh depth of the historical resources;
[0042] S204, determining resource satisfaction information of the current user at the current refresh depth according to the playback time of the reference resource by the current user and the total playback time of the reference resource;
[0043] S205, screening the candidate resources according to the resource satisfaction information and the similarity, and determining the to-be-recommended resources for the current user at the current refresh depth based on the screened candidate resources based on the big model.
[0044] Among them, historical refresh depth, playback duration and total duration belong to playback attribute information of historical resources. In the process of screening each candidate resource, reference resources associated with the current refresh depth can be selected from the historical resources according to the current refresh depth and the historical refresh depth of the historical resources. Exemplarily, historical resources with the same historical refresh depth as the current refresh depth can be used as reference resources for the current refresh depth; the refresh depths can also be grouped to obtain multiple refresh depth intervals, and historical resources belonging to the same refresh depth interval as the current refresh depth can be used as reference resources for the current refresh depth. By selecting reference resources associated with the current refresh depth from historical resources, the matching degree between the reference resources and the current refresh depth is improved compared to using all historical resources as reference resources.
[0045] Exemplarily, the current user's satisfaction with the reference resource can be determined based on the current user's playback time of the reference resource and the total time of the reference resource, and the resource satisfaction information of the current refresh depth can be determined based on the satisfaction with the reference resource. By using the playback time and total time of the reference resource to determine the resource satisfaction information of the current refresh depth, the matching degree between the resource satisfaction information and the current refresh depth is also improved, that is, the accuracy of the resource satisfaction information is improved.
[0046] In an optional implementation, determining the resource satisfaction information of the current user at the current refresh depth based on the playback time of the reference resource by the current user and the total duration of the reference resource includes: determining the satisfaction of the current user with the reference resource at the current refresh depth based on the playback time of the reference resource by the current user and the total duration of the reference resource; selecting a first resource that the current user is satisfied with and a second resource that the current user is not satisfied with from the reference resources based on the satisfaction; and determining the resource satisfaction information based on the first resource and the second resource.
[0047] Exemplarily, the current user's satisfaction with the reference resources at the current refresh depth can be determined in combination with the playback time and total time of the reference resources; the first value of reference resources with relatively high satisfaction are used as the first resources that the current user is satisfied with; the second value of reference resources with relatively low satisfaction are used as the second resources that the current user is satisfied with. Among them, the first value and the second value are preset, for example, they can both be 20. Determine the resource satisfaction information in combination with each first resource and each second resource. For example, the gap between the first resource and the second resource in attributes such as content and function can be analyzed to determine the resource satisfaction information. By determining the current user's satisfaction with each reference resource, and selecting each first resource that the current user is satisfied with and each second resource that the current user is not satisfied with from the reference resources according to the satisfaction; determining the resource satisfaction information in combination with the first resource that the current user is satisfied with and the second resource that the current user is not satisfied with, the resource content difference between the first resource and the second resource itself can be fully considered, thereby helping to improve the accuracy and quality of the resource satisfaction information.
[0048] In an optional implementation, determining the current user's satisfaction with the reference resource at the current refresh depth based on the playback time of the reference resource by the current user and the total playback time of the reference resource includes: determining the playback completion of the reference resource based on the reference playback time of the current user and the total playback time of the reference resource; obtaining the maximum playback time of a single resource from the reference playback time, and determining the contribution of the reference resource based on the reference playback time and the maximum playback time; and determining the satisfaction based on the playback completion and the contribution.
[0049] Exemplarily, the reference playback duration of the reference resource may be divided by the total duration of the reference resource to obtain the playback completion of the reference resource; the playback completion of the reference resource may also be determined by the following formula:
[0050] finish_rate=min(1.0, play_duration / sv_duration)
[0051] Among them, play_duration, sv_duration and finish_rate are the reference playback duration, total duration and playback completion rate respectively, and min() is used to obtain the minimum value.
[0052] Moreover, the maximum playback time of a single resource is obtained from each reference playback time, and the reference playback time can be divided by the maximum playback time to obtain the contribution of the reference resource; the satisfaction of the reference resource can be determined in combination with the playback completion degree and the contribution of the reference resource. For example, the satisfaction degree can be positively correlated with the playback completion degree, and the satisfaction degree can change rapidly with the increase of contribution and approach a fixed value, which can be 1.
[0053] Each user has different consumption habits. Some users may prefer to scroll quickly and rarely watch resources for a long time, while some users may watch most resources for a long time. By taking the maximum playback time of the current user as the benchmark for the current user and determining the contribution of the reference resource based on the reference playback time and maximum playback time of the reference resource, the matching degree between the contribution and the current user can be improved, and the accuracy of the contribution can be improved. By combining the playback completion and contribution of the reference resource to determine the satisfaction of the reference resource, the influence of user consumption habits on satisfaction can be eliminated, thereby improving the accuracy of satisfaction.
[0054] In an optional implementation manner, determining the satisfaction level according to the playback completion level and the contribution level includes determining the satisfaction level by using the following formula:
[0055]
[0056] Among them, satisfy_score is satisfaction, finish_rate is play completion, play_rate is contribution, and a is a constant. For example, the value of a can be 8. Through the above processing, satisfaction and play completion are positively correlated, and as the contribution increases, the corresponding satisfaction increases rapidly until it stabilizes.
[0057] The technical solution provided by the embodiments of the present disclosure determines the playback completion and contribution of the reference resources by adopting the playback duration and total duration of the reference resources, and then determines the satisfaction of the reference resources; and selects the first resources that the current user is satisfied with and the second resources that the current user is not satisfied with from the reference resources according to the satisfaction; and determines the resource satisfaction information by combining the first resources that the current user is satisfied with and the second resources that the current user is not satisfied with, which can fully consider the resource content differences between the first resources and the second resources themselves, thereby helping to improve the accuracy and quality of the resource satisfaction information.
[0058] In an optional implementation, the resource satisfaction information is determined based on the first resource and the second resource, including: averaging the vector representations of the first resources to obtain a first resource vector, and averaging the vector representations of the second resources to obtain a second resource vector; determining a first correlation between the vector representation of the candidate resource and the first resource vector, and determining a second correlation between the vector representation of the candidate resource and the second resource vector; subtracting the second correlation from the first correlation to obtain a resource satisfaction coefficient of the current refresh depth; the resource satisfaction information is the resource satisfaction coefficient.
[0059] Still taking the acquisition of 20 first resources and 20 second resources as an example, illustratively, the vector representations (embeddings) of the 20 first resources can be averaged to obtain the first resource vector, so that the first resource vector matches the interest point of the current user; the vector representations of the 20 second resources can be averaged to obtain the second resource vector, so that the second resource vector matches the aversion point of the current user. And, determine the first correlation between the vector representation of the candidate resource and the first resource vector, determine the second correlation between the vector representation of the candidate resource and the second resource vector, and subtract the second correlation from the first correlation to obtain the resource satisfaction coefficient of the current refresh depth, and use the resource satisfaction coefficient as the resource satisfaction information. By subtracting the second correlation between the candidate resource and the second resource vector corresponding to the aversion point from the first correlation between the candidate resource and the first resource vector matching the interest point, the resource satisfaction coefficient is obtained, that is, the resource satisfaction coefficient can take into account the interest point and the aversion point of the current user, so that the recommended resource determined according to the resource satisfaction coefficient not only meets the interest point of the current user, but also avoids the aversion point of the current user, further improving the matching degree between the recommended resource and the current user, and improving the user experience.
[0060] In an optional implementation, screening the candidate resources according to the resource satisfaction information and the similarity includes: multiplying the resource satisfaction coefficient by the similarity to obtain a new similarity of the candidate resources; and screening the candidate resources according to the new similarity.
[0061] Exemplarily, candidate resources with relatively low new similarity can be filtered out, and only candidate resources with relatively high new similarity can be retained. By multiplying the similarity of the candidate resource and the resource satisfaction coefficient as the new similarity of the candidate resource, when the resource satisfaction coefficient is much smaller than the similarity and non-zero, the influence of the resource satisfaction coefficient on the similarity can be retained, thereby reflecting the matching degree of the candidate resource.
[0062] Figure 3 is a flowchart of another method for determining resources to be recommended based on a large model according to an embodiment of the present disclosure. Figure 3 The method for determining resources to be recommended based on a large model in this embodiment may include:
[0063] S301, obtaining the current user's behavior preference and current refresh depth;
[0064] S302, determining the similarity between a candidate resource and the candidate resource according to the behavior preference and the initial resource;
[0065] S303, obtaining the historical resources played by the current user, and determining the number of historical resources played by the current user at each historical refresh depth according to the historical refresh depth of the historical resources;
[0066] S304, dividing each of the historical refresh depths by using the number of historical resources to obtain at least two refresh depth intervals;
[0067] S305, determining the current refresh depth interval to which the current refresh depth belongs from the at least two refresh depth intervals, and using historical resources whose historical refresh depths belong to the current refresh depth interval as reference resources;
[0068] S306, determining resource satisfaction information of the current user at the current refresh depth according to the playback time of the reference resource by the current user and the total playback time of the reference resource;
[0069] S307, screening the candidate resources according to the resource satisfaction information and the similarity, and determining the recommended resources for the current user at the current refresh depth based on the screened candidate resources based on the big model.
[0070] Among them, historical refresh depth, playback duration and total duration belong to the playback attribute information of historical resources.
[0071] In the process of screening each candidate resource, the historical resources played by the current user and the historical refresh depth of each historical resource in the corresponding historical playback session are obtained; for each historical refresh depth, the number of historical resources played by the current user at each historical refresh depth is obtained, that is, the number of historical resources belonging to the historical refresh depth. For example, the current user played 5 historical resources in the first historical playback session, 4 historical resources in the second historical playback session, and 7 historical resources in the third historical playback session. The number of historical resources at the first historical refresh depth, the second historical refresh depth, the third historical refresh depth, and the fourth historical refresh depth are all 3, the number of historical resources at the fifth historical refresh depth is 2, and the number of historical resources at the sixth historical refresh depth and the seventh historical refresh depth are all 1.
[0072] Exemplarily, the historical resources played by the current user in different historical playback sessions and the historical refresh depth of each historical resource are obtained; for each historical refresh depth, the number of historical playback sessions including the historical refresh depth is obtained as the number of historical resources for the historical refresh depth. Each historical refresh depth is divided according to the number of historical resources at each historical refresh depth to obtain at least two refresh depth intervals; the refresh depth interval to which the current refresh depth belongs is taken as the current refresh depth interval, and each historical resource whose historical refresh depth belongs to the current refresh depth interval is taken as a reference resource. By dividing each historical refresh depth by the number of historical resources to obtain at least two refresh depth intervals, taking the refresh depth interval to which the current refresh depth belongs as the current depth interval, and taking the historical resources whose historical refresh depth belongs to the current refresh depth interval as reference resources, the number of reference resources can be enriched, thereby improving the quality of subsequent processing based on the playback time of the reference resources and the total time of the reference resources.
[0073] In an optional implementation, the deviation between the historical resource quantities in different refresh depth intervals is smaller than a preset deviation threshold.
[0074] Exemplarily, in the process of dividing each historical refresh depth to obtain at least two refresh depth intervals, the number of historical resources in each refresh depth interval is made approximately equal. Since the shallower the historical refresh depth, the more historical resources the corresponding number is relatively, and the deeper the historical refresh depth, the fewer historical resources the corresponding number is relatively. In the process of dividing the historical refresh depth, the number of historical resources corresponding to each refresh depth interval is controlled to be approximately equal, so that the number of historical refresh depths in the relatively shallow refresh depth interval is less than the number of historical refresh depths in the relatively deep refresh depth interval. For example, the first refresh depth interval may include 4 historical refresh depths of 1-4, the second refresh depth interval may include 6 historical refresh depths of 5-10, and the third refresh depth interval may include 11-20, a total of 10 historical refresh depths. By controlling the number of historical resources in each refresh depth interval to be approximately equal, the number of reference resources corresponding to each refresh depth interval is made approximately equal, thereby balancing the impact of reference resources on resource satisfaction information in each refresh depth interval, and improving the generalization of resource depth information.
[0075] The technical solution provided by the embodiment of the present disclosure divides each historical refresh depth into at least two refresh depth intervals by using the number of historical resources, takes the refresh depth interval to which the current refresh depth belongs as the current depth interval, and takes the historical resources whose historical refresh depths belong to the current refresh depth interval as reference resources, thereby enriching the number of reference resources, thereby improving the quality of subsequent processing based on the playback time of the reference resources and the total time of the reference resources.
[0076] Figure 41 is a schematic diagram of a structure of a device for determining resources to be recommended based on a large model according to an embodiment of the present disclosure. The device is suitable for pushing resources to users. The device can be implemented in software and / or hardware and can be integrated into electronic devices. Figure 4 As shown, the apparatus 400 for determining resources to be recommended based on a large model in this embodiment may include:
[0077] The current acquisition module 410 is used to acquire the current user's behavior preference and current refresh depth;
[0078] A candidate resource module 420, configured to determine the similarity between a candidate resource and the candidate resource according to the behavior preference and the initial resource;
[0079] The resource satisfaction module 430 is used to determine the resource satisfaction information of the current user at the current refresh depth according to the current refresh depth, the historical resources played by the current user and the playback attribute information of the historical resources; wherein the playback attribute information at least includes the historical refresh depth;
[0080] The resource module 440 to be recommended is used to screen the candidate resources according to the resource satisfaction information and the similarity, and determine the resource to be recommended for the current user at the current refresh depth based on the screened candidate resources based on the big model.
[0081] In an optional implementation, the resource satisfaction module 430 includes:
[0082] A reference resource submodule, configured to determine a reference resource associated with the current refresh depth from the historical resources according to the current refresh depth and the historical refresh depth of the historical resources;
[0083] The resource satisfaction submodule is used to determine the resource satisfaction information of the current user at the current refresh depth based on the playback time of the reference resource by the current user and the total playback time of the reference resource; wherein the playback time and the total playback time belong to the playback attribute information.
[0084] In an optional implementation, the resource satisfaction submodule includes:
[0085] A satisfaction unit, configured to determine the satisfaction of the current user with the reference resource at the current refresh depth according to the playback duration of the reference resource by the current user and the total duration of the reference resource;
[0086] A resource selection unit, configured to select, from the reference resources, a first resource that the current user is satisfied with and a second resource that the current user is not satisfied with according to the satisfaction level;
[0087] A resource satisfaction unit is used to determine the resource satisfaction information according to the first resource and the second resource.
[0088] In an optional implementation, the satisfaction unit includes:
[0089] A completion subunit, configured to determine the completion of playback of the reference resource according to the reference playback duration of the reference resource by the current user and the total duration of the reference resource;
[0090] A contribution subunit, used to obtain a maximum playback duration of a single resource from the reference playback duration, and determine a contribution of the reference resource according to the reference playback duration and the maximum playback duration;
[0091] The satisfaction subunit is used to determine the satisfaction level according to the playback completion level and the contribution level.
[0092] In an optional implementation manner, the satisfaction subunit is specifically used to determine the satisfaction through the following formula:
[0093]
[0094] Among them, satisfy_score is the satisfaction level, finish_rate is the playback completion rate, play_rate is the contribution rate, and a is a constant.
[0095] In an optional implementation, the resource satisfaction unit includes:
[0096] a resource vector subunit, configured to average the vector representations of the first resources to obtain a first resource vector, and average the vector representations of the second resources to obtain a second resource vector;
[0097] a resource relevance subunit, configured to determine a first relevance between the vector representation of the candidate resource and the first resource vector, and to determine a second relevance between the vector representation of the candidate resource and the second resource vector;
[0098] The satisfaction coefficient subunit is used to subtract the second relevance from the first relevance to obtain the resource satisfaction coefficient of the current refresh depth; the resource satisfaction information is the resource satisfaction coefficient.
[0099] In an optional implementation, the to-be-recommended resource module 440 includes:
[0100] A new similarity submodule, used for multiplying the resource satisfaction coefficient by the similarity to obtain a new similarity of the candidate resource;
[0101] The resource screening submodule is used to screen the candidate resources according to the new similarity.
[0102] In an optional implementation, the reference resource submodule includes:
[0103] A resource quantity unit, used to determine the quantity of historical resources played by the current user at each historical refresh depth according to the historical refresh depth of the historical resources;
[0104] A refresh depth division unit, used to divide each of the historical refresh depths using the number of historical resources to obtain at least two refresh depth intervals;
[0105] A reference resource unit is used to determine the current refresh depth interval to which the current refresh depth belongs from the at least two refresh depth intervals; wherein the reference resource is a historical resource whose historical refresh depth belongs to the current refresh depth interval.
[0106] In an optional implementation, the deviation between the historical resource quantities in different refresh depth intervals is smaller than a preset deviation threshold.
[0107] The device for determining resources to be recommended based on a large model provided in an embodiment of the present invention can execute the method for determining resources to be recommended based on a large model provided in any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.
[0108] In the technical solution disclosed herein, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.
[0109] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.
[0110] Figure 5 The block diagram of an electronic device is used to implement the method for determining resources to be recommended based on a large model according to an embodiment of the present disclosure. Figure 5 A schematic block diagram of an example electronic device 500 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.
[0111] like Figure 5 As shown, the electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.
[0112] Multiple components in the electronic device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the electronic device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0113] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 501 performs the various methods and processes described above, such as a method for determining resources to be recommended based on a large model. For example, in some embodiments, the method for determining resources to be recommended based on a large model may be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 500 via the ROM 502 and / or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the method for determining resources to be recommended based on a large model described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured in any other appropriate manner (for example, by means of firmware) to execute the method for determining resources to be recommended based on a large model.
[0114] Figure 5A schematic block diagram of an example electronic device 500 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.
[0115] like Figure 5 As shown, the electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.
[0116] Multiple components in the electronic device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the electronic device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0117] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 501 performs the various methods and processes described above, such as a method for determining resources to be recommended based on a large model. For example, in some embodiments, the method for determining resources to be recommended based on a large model may be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 500 via the ROM 502 and / or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the method for determining resources to be recommended based on a large model described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured in any other appropriate manner (for example, by means of firmware) to execute the method for determining resources to be recommended based on a large model.
[0118] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0119] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.
[0120] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0121] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, audio input, or tactile input).
[0122] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web player through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.
[0123] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.
[0124] Artificial intelligence is a discipline that studies how to use computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, planning, etc.). It includes both hardware-level and software-level technologies. Artificial intelligence hardware technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, and big data processing; artificial intelligence software technologies mainly include computer vision technology, audio recognition technology, natural language processing technology, machine learning / deep learning technology, big data processing technology, knowledge graph technology, and other major directions.
[0125] Cloud computing refers to a technology system that uses network access to elastically scalable shared physical or virtual resource pools. Resources can include servers, operating systems, networks, software, applications, and storage devices, and can be deployed and managed on demand and in a self-service manner. Cloud computing technology can provide efficient and powerful data processing capabilities for technical applications such as artificial intelligence and blockchain, as well as model training.
[0126] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.
[0127] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.
Claims
1. A method for determining resources to be recommended based on a large model, comprising: Get the current user's behavior preferences and current refresh depth; Determining, based on the behavior preference and the initial resource, a similarity between the candidate resource and the candidate resource; Determine the resource satisfaction information of the current user at the current refresh depth according to the current refresh depth, the historical resources played by the current user and the playback attribute information of the historical resources; wherein the playback attribute information at least includes the historical refresh depth; The candidate resources are screened according to the resource satisfaction information and the similarity, and the recommended resources for the current user at the current refresh depth are determined based on the screened candidate resources based on the big model.
2. The method according to claim 1, wherein: The determining, according to the current refresh depth, the historical resources played by the current user, and the playback attribute information of the historical resources, the resource satisfaction information of the current user at the current refresh depth includes: Determining, from the historical resources, a reference resource associated with the current refresh depth according to the current refresh depth and the historical refresh depth of the historical resources; The resource satisfaction information of the current user at the current refresh depth is determined based on the playback duration of the reference resource by the current user and the total duration of the reference resource; wherein the playback duration and the total duration belong to the playback attribute information.
3. The method according to claim 2, wherein: Determining the resource satisfaction information of the current user at the current refresh depth according to the playback time of the reference resource by the current user and the total playback time of the reference resource includes: Determining the satisfaction of the current user with the reference resource at the current refresh depth according to the playback duration of the reference resource by the current user and the total duration of the reference resource; According to the satisfaction level, selecting a first resource that the current user is satisfied with and a second resource that the current user is not satisfied with from the reference resources; The resource satisfaction information is determined according to the first resource and the second resource.
4. The method according to claim 3, wherein: The determining, according to the playback duration of the reference resource by the current user and the total duration of the reference resource, the satisfaction of the current user with the reference resource at the current refresh depth includes: Determining the playback completion degree of the reference resource according to the reference playback duration of the reference resource by the current user and the total playback duration of the reference resource; Acquire a maximum playback duration of a single resource from the reference playback duration, and determine a contribution of the reference resource according to the reference playback duration and the maximum playback duration; The satisfaction level is determined according to the playback completion level and the contribution level.
5. The method according to claim 4, wherein: Determining the satisfaction level according to the playback completion level and the contribution level includes determining the satisfaction level by the following formula: Among them, satisfy_score is the satisfaction level, finish_rate is the playback completion rate, play_rate is the contribution rate, and a is a constant.
6. The method according to any one of claims 3 to 5, wherein: The determining the resource satisfaction information according to the first resource and the second resource includes: Averaging the vector representations of the first resources to obtain a first resource vector, and averaging the vector representations of the second resources to obtain a second resource vector; Determining a first correlation between the vector representation of the candidate resource and the first resource vector, and determining a second correlation between the vector representation of the candidate resource and the second resource vector; The resource satisfaction coefficient of the current refresh depth is obtained by subtracting the second correlation from the first correlation; the resource satisfaction information is the resource satisfaction coefficient.
7. The method according to claim 6, wherein: The screening of the candidate resources according to the resource satisfaction information and the similarity includes: Multiplying the resource satisfaction coefficient by the similarity to obtain a new similarity of the candidate resource; The candidate resources are screened according to the new similarity.
8. The method according to claim 2, wherein: The determining, from the historical resources according to the current refresh depth and the historical refresh depth of the historical resources, a reference resource associated with the current refresh depth comprises: Determining the number of historical resources played by the current user at each historical refresh depth according to the historical refresh depth of the historical resources; Dividing each of the historical refresh depths using the number of historical resources to obtain at least two refresh depth intervals; Determine the current refresh depth interval to which the current refresh depth belongs from the at least two refresh depth intervals; wherein the reference resource is a historical resource whose historical refresh depth belongs to the current refresh depth interval.
9. The method according to claim 8, wherein: The deviation between the historical resource quantities in different refresh depth intervals is less than a preset deviation threshold.
10. A device for determining resources to be recommended based on a large model, comprising: The current acquisition module is used to obtain the current user's behavior preferences and current refresh depth; A candidate resource module, used to determine the similarity between a candidate resource and the candidate resource according to the behavior preference and the initial resource; A resource satisfaction module, used to determine the resource satisfaction information of the current user at the current refresh depth according to the current refresh depth, the historical resources played by the current user and the playback attribute information of the historical resources; wherein the playback attribute information at least includes the historical refresh depth; The resource module to be recommended is used to screen the candidate resources according to the resource satisfaction information and the similarity, and determine the resource to be recommended for the current user at the current refresh depth based on the screened candidate resources based on the big model.
11. The device according to claim 10, wherein: The resource satisfaction module includes: A reference resource submodule, configured to determine a reference resource associated with the current refresh depth from the historical resources according to the current refresh depth and the historical refresh depth of the historical resources; The resource satisfaction submodule is used to determine the resource satisfaction information of the current user at the current refresh depth based on the playback time of the reference resource by the current user and the total playback time of the reference resource; wherein the playback time and the total playback time belong to the playback attribute information.
12. The device according to claim 11, wherein The resource satisfaction submodule includes: A satisfaction unit, configured to determine the satisfaction of the current user with the reference resource at the current refresh depth according to the playback duration of the reference resource by the current user and the total duration of the reference resource; A resource selection unit, configured to select, from the reference resources, a first resource that the current user is satisfied with and a second resource that the current user is not satisfied with according to the satisfaction level; A resource satisfaction unit is used to determine the resource satisfaction information according to the first resource and the second resource.
13. The device according to claim 12, wherein: The satisfaction unit includes: A completion subunit, configured to determine the completion of playback of the reference resource according to the reference playback duration of the reference resource by the current user and the total duration of the reference resource; A contribution subunit, used to obtain a maximum playback duration of a single resource from the reference playback duration, and determine a contribution of the reference resource according to the reference playback duration and the maximum playback duration; The satisfaction subunit is used to determine the satisfaction level according to the playback completion level and the contribution level.
14. The device according to claim 13, wherein: The satisfaction subunit is specifically used to determine the satisfaction through the following formula: Among them, satisfy_score is the satisfaction level, finish_rate is the playback completion rate, play_rate is the contribution rate, and a is a constant.
15. The device according to any one of claims 12 to 14, wherein: The resource satisfaction unit includes: a resource vector subunit, configured to average the vector representations of the first resources to obtain a first resource vector, and average the vector representations of the second resources to obtain a second resource vector; a resource relevance subunit, configured to determine a first relevance between the vector representation of the candidate resource and the first resource vector, and to determine a second relevance between the vector representation of the candidate resource and the second resource vector; The satisfaction coefficient subunit is used to subtract the second relevance from the first relevance to obtain the resource satisfaction coefficient of the current refresh depth; the resource satisfaction information is the resource satisfaction coefficient.
16. The device according to claim 15, wherein: The resource module to be recommended includes: A new similarity submodule, used for multiplying the resource satisfaction coefficient by the similarity to obtain a new similarity of the candidate resource; The resource screening submodule is used to screen the candidate resources according to the new similarity.
17. The device according to claim 11, wherein: The reference resource submodule includes: A resource quantity unit, used to determine the quantity of historical resources played by the current user at each historical refresh depth according to the historical refresh depth of the historical resources; A refresh depth division unit, used to divide each of the historical refresh depths using the number of historical resources to obtain at least two refresh depth intervals; A reference resource unit is used to determine the current refresh depth interval to which the current refresh depth belongs from the at least two refresh depth intervals; wherein the reference resource is a historical resource whose historical refresh depth belongs to the current refresh depth interval.
18. The device according to claim 17, wherein: The deviation between the historical resource quantities in different refresh depth intervals is less than a preset deviation threshold.
19. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 9.
20. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to make a computer execute the method according to any one of claims 1-9.
21. A computer program product comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 9.
Citation Information
Patent Citations
Method and device for refreshing news list
CN104239587A
Network resource recommendation method and apparatus, electronic device, server and storage medium
CN108304441A
Resource data processing method and device, electronic equipment and medium
CN111563202A
Multimedia resource display method, device and system and storage medium
CN112131410A
Resource recommendation method and device, equipment and storage medium
CN116701753A
