Recommended content generation method, apparatus, device, storage medium, and product

Abstract

The application discloses a recommendation content generation method and device, equipment, storage medium and product, relates to the technical field of artificial intelligence, and includes the following steps: in response to demand information triggered in a browser, splitting the demand information; selecting target artificial intelligence models respectively meeting each sub-demand information from a plurality of artificial intelligence models accessed to the browser; determining a cooperative operation mode between the target artificial intelligence models, and generating recommendation content corresponding to the demand information based on the cooperative operation mode through the target artificial intelligence models; and in the above manner, the triggered demand information is split into each sub-demand information, and after the target artificial intelligence models respectively meeting each sub-demand information are selected, the target artificial intelligence models are utilized to generate recommendation content meeting the demand information based on the cooperative operation mode between the target artificial intelligence models, so that the accuracy of generating the recommendation content can be effectively improved, the experience of a user is improved, and diversified demands of the user are met.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to methods, apparatus, devices, storage media, and products for generating recommendation content. Background Technology

[0002] As artificial intelligence technology develops, user needs are becoming increasingly complex and diverse. Existing single AI models are unable to meet these needs. For example, a user's application requirement is to translate articles, which involves two steps: searching for articles and translating the content. However, a single AI model only has a search function and cannot translate. In this case, the generated recommendation content would be: "Cannot be translated," along with the specific search content, which does not meet the user's needs. Therefore, the accuracy of the recommended content generated in this way is low.

[0003] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main objective of this application is to provide a method, apparatus, device, storage medium, and product for generating recommended content, aiming to solve the technical problem of low accuracy of recommended content in the prior art.

[0005] To achieve the above objectives, this application proposes a method for generating recommended content, the method comprising:

[0006] In response to the request information triggered in the browser, the request information is split into sub-request information;

[0007] Select a target artificial intelligence model from the various artificial intelligence models accessed by the browser that respectively meet the sub-requirement information;

[0008] Determine the collaborative operation mode among the target artificial intelligence models, and generate recommended content corresponding to the demand information based on the collaborative operation mode using the target artificial intelligence models.

[0009] In one embodiment, the step of selecting a target artificial intelligence model from multiple artificial intelligence models accessing the browser that respectively meet the sub-requirement information includes:

[0010] Obtain functional description information of various artificial intelligence models accessed by the browser;

[0011] Determine the functions required to satisfy each of the sub-requirements;

[0012] The required functions are matched with the function description information respectively;

[0013] Based on the matching results, a target artificial intelligence model that satisfies each sub-requirement information is selected from the various artificial intelligence models.

[0014] In one embodiment, the step of determining the collaborative operation mode between the target artificial intelligence models includes:

[0015] The target artificial intelligence models are grouped to obtain a combination of target artificial intelligence models;

[0016] Dependency detection is performed on each of the aforementioned target AI model combinations;

[0017] The collaborative operation mode between the target artificial intelligence models is determined based on the dependency detection results.

[0018] In one embodiment, the step of determining the collaborative operation mode between the target artificial intelligence models based on the dependency detection results includes:

[0019] Obtain the operational description information of each of the target artificial intelligence models;

[0020] Based on the operational description information, obtain the resource information required for the normal operation of each of the target artificial intelligence models;

[0021] The collaborative operation mode among the target artificial intelligence models is determined based on the dependency detection results and the required resource information.

[0022] In one embodiment, the step of determining the collaborative operation mode among the target artificial intelligence models based on the dependency detection results and the required resource information includes:

[0023] Obtain information on remaining system resources;

[0024] The collaborative running time of each target artificial intelligence model is determined based on the dependency detection results, the required resource information, and the remaining system resource information.

[0025] The collaborative operation mode among the target artificial intelligence models is determined based on each of the collaborative operation moments.

[0026] In one embodiment, the step of determining the collaborative execution time of each target artificial intelligence model based on the dependency detection result, the required resource information, and the remaining system resource information includes:

[0027] Based on the dependency detection results, target AI models with and without dependency relationships are obtained;

[0028] Select a target AI model that runs synchronously with the target AI model that does not have a dependency relationship from among the target AI models that have a dependency relationship;

[0029] Determine the total resource information required for the target artificial intelligence model to run synchronously based on the resource information described above.

[0030] The total resource information, the remaining system resource information, and each of the required resource information are input into the collaborative operation time prediction model to obtain the collaborative operation time of each target artificial intelligence model output by the collaborative operation time prediction model.

[0031] In one embodiment, before the step of selecting a target artificial intelligence model that satisfies each sub-requirement information from multiple artificial intelligence models accessing the browser, the method further includes:

[0032] A standard multi-model access framework is generated based on the unified strategy for interface protocols and the unified strategy for data exchange.

[0033] The AI ​​model to be accessed is connected to the browser based on the standard multi-model access framework.

[0034] In one embodiment, the step of connecting the AI ​​model to be accessed to the browser based on the standard multi-model access framework includes:

[0035] Obtain model feature information for each artificial intelligence model supported by the browser;

[0036] Based on the model feature information, generate artificial intelligence model access conditions;

[0037] Obtain the attribute information of the artificial intelligence model to be connected;

[0038] Based on the standard multi-model access framework, the artificial intelligence model to be accessed is accessed to the browser according to the attribute information and the access conditions of the artificial intelligence model.

[0039] In one embodiment, the step of connecting the artificial intelligence model to be accessed to the browser based on the standard multi-model access framework, according to the attribute information and the artificial intelligence model access conditions, includes:

[0040] Based on the attribute information, determine whether there are any artificial intelligence models among the artificial intelligence models to be accessed that do not meet the access conditions for artificial intelligence models;

[0041] If so, then data conversion is performed on the artificial intelligence models that do not meet the artificial intelligence model access conditions based on the standard multi-model access framework;

[0042] The converted AI model is then connected to the browser.

[0043] In one embodiment, the step of generating recommended content corresponding to the demand information based on the collaborative operation mode using the target artificial intelligence model includes:

[0044] Based on the collaborative operation mode, control the collaborative operation of each target artificial intelligence model and obtain the sub-recommendation content output by each target artificial intelligence model;

[0045] Based on each of the sub-recommendation contents, generate recommended content corresponding to the demand information.

[0046] In one embodiment, after the step of responding to the request information triggered in the browser, the method further includes:

[0047] The demand information is subjected to feature extraction to obtain demand features;

[0048] When the required feature is an application display feature, search for applications integrated into the browser;

[0049] An application list is generated based on each application and its attribute information, and the application list is then displayed.

[0050] In one embodiment, after the step of displaying the application list, the method further includes:

[0051] In response to application update information triggered in the browser, an application update data packet is determined based on the application update information;

[0052] Extract the header of the application update data packet;

[0053] The application to be updated in the application list is determined based on the header;

[0054] The application to be updated is updated according to the application update data package.

[0055] Furthermore, to achieve the above objectives, this application also proposes a recommended content generation apparatus, which includes:

[0056] The splitting module is used to split the demand information triggered in the browser to obtain sub-demand information in response to the demand information.

[0057] The selection module is used to select a target artificial intelligence model that meets each sub-requirement information from a variety of artificial intelligence models accessed by the browser;

[0058] The generation module is used to determine the collaborative operation mode between the target artificial intelligence models, and generate recommended content corresponding to the demand information based on the collaborative operation mode by the target artificial intelligence models.

[0059] In one embodiment, the selection module is further configured to acquire functional description information of multiple artificial intelligence models accessing the browser; determine the functions required to satisfy each sub-requirement information; match the required functions with the functional description information; and select a target artificial intelligence model from the multiple artificial intelligence models that satisfies each sub-requirement information based on the matching results.

[0060] In one embodiment, the generation module is further configured to group the target artificial intelligence models to obtain target artificial intelligence model combinations; perform dependency detection on each target artificial intelligence model combination; and determine the collaborative operation mode between the target artificial intelligence models based on the dependency detection results.

[0061] In one embodiment, the generation module is further configured to obtain runtime description information of each of the target artificial intelligence models; obtain resource information required for the normal operation of each of the target artificial intelligence models based on the runtime description information; and determine the collaborative operation mode between the target artificial intelligence models based on the dependency detection results and the required resource information.

[0062] In one embodiment, the generation module is further configured to acquire system remaining resource information; determine the collaborative running time of each target artificial intelligence model based on the dependency detection result, the required resource information, and the system remaining resource information; and determine the collaborative running mode between the target artificial intelligence models based on the collaborative running time.

[0063] In addition, to achieve the above objectives, this application also proposes a recommended content generation device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the recommended content generation method as described above.

[0064] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and which, when executed by a processor, implements the steps of the recommendation content generation method described above.

[0065] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the recommended content generation method described above.

[0066] One or more technical solutions proposed in this application have at least the following technical effects: by responding to demand information triggered in a browser, the demand information is split into sub-demand information; target artificial intelligence models that satisfy each sub-demand information are selected from multiple artificial intelligence models accessed by the browser; the collaborative operation mode between the target artificial intelligence models is determined, and recommended content corresponding to the demand information is generated by the target artificial intelligence models based on the collaborative operation mode; by splitting the triggered demand information into sub-demand information in the above manner, and after selecting target artificial intelligence models that satisfy each sub-demand information, recommended content that satisfies the demand information is generated based on the collaborative operation mode between the target artificial intelligence models, thereby effectively improving the accuracy of the generated recommended content, thereby improving the user experience and meeting the complex and diverse needs of users. Attached Figure Description

[0067] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0068] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0069] Figure 1 A flowchart illustrating the method for generating recommended content in this application, provided in Embodiment 1.

[0070] Figure 2 A flowchart illustrating the second embodiment of the method for generating recommended content in this application;

[0071] Figure 3 This is a schematic diagram of the module structure of the recommended content generation device in the embodiments of this application;

[0072] Figure 4 This is a schematic diagram of the device structure of the hardware operating environment involved in the recommended content generation method in the embodiments of this application.

[0073] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0074] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device or recommendation content generation device capable of performing the above functions. The following description uses a recommendation content generation device as an example to illustrate this embodiment and the subsequent embodiments.

[0075] Based on this, embodiments of this application provide a method for generating recommended content, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the method for generating recommended content in this application.

[0076] In this embodiment, the recommended content generation method includes steps S10 to S30:

[0077] Step S10: In response to the demand information triggered in the browser, the demand information is split to obtain sub-demand information.

[0078] It should be noted that the browser in this embodiment can be an artificial intelligence (AI) browser that is equipped with a hybrid artificial intelligence model, intelligent scheduling algorithm, etc. This browser can also be responsible for interface display and user interaction, such as generating and displaying recommended content that matches the demand information, displaying an application list, and having a user-friendly interface and simple operation.

[0079] It should be understood that demand information refers to the information that users enter in the input box set in the browser according to their actual needs. In order to generate recommended content that meets the demand information, the demand information is broken down into multiple sub-demand information. For example, if the triggered demand information is to create a PPT, then the sub-demand information is to search for materials, construct a framework, and generate the final PPT. The searched materials include, but are not limited to, images and text.

[0080] Furthermore, to better meet user needs and improve user experience, after the step of responding to demand information triggered in the browser, the method further includes: extracting features from the demand information to obtain demand features; when the demand features are application display features, searching for applications integrated in the browser; generating an application list based on each application and its attribute information, and displaying the application list.

[0081] It should be understood that when the demand characteristics corresponding to the demand information are determined to be application display characteristics, it indicates that the user wants to view applications integrated into the browser. In this case, the system searches for applications integrated into the browser and generates an application list based on the application's attribute information, such as search applications or translation applications. Then, the applications in the application list can be sorted, and the sorted application list can be displayed. The metrics used for sorting can be determined comprehensively based on user feedback data and application performance.

[0082] It should be noted that the browser will also have a pre-set unified management application module, which can be a module that integrates download, feedback and update functions. In the unified management application module, users can quickly search for applications that suit their needs.

[0083] Furthermore, to ensure timely and accurate application updates, after displaying the application list, the method further includes: responding to application update information triggered in the browser, determining an application update data packet based on the application update information; extracting the header of the application update data packet; determining the applications to be updated in the application list based on the header; and updating the applications to be updated based on the application update data packet.

[0084] It is understood that application update information refers to information about updating applications integrated into the browser, such as updating the application version from 18.0 to 18.1. This embodiment provides developers with a backend interface for uploading application update information and managing applications. After extracting the header of the application update data package, the application to be updated in the application list can be determined based on the header. For example, the application to be updated may be the search application in the application list. Then, the application to be updated is updated according to the application update data package, automatically completing the application configuration and release, reducing the difficulty of user operation.

[0085] Step S20: Select a target artificial intelligence model from the various artificial intelligence models accessed by the browser that respectively meet the sub-requirement information.

[0086] It is understandable that "multiple AI models" refers to AI models that have been successfully and seamlessly integrated into the browser. These multiple AI models include, but are not limited to, large-scale enterprise private models and large-scale open internet models. When integrated, they carry functional description information of various AI models. This functional description information can be carried in the form of tags. For example, AI model A is used to search for materials, satisfying sub-requirement information a; AI model B is used to construct a framework, satisfying sub-requirement information b; and AI model C is used to generate PPTs, satisfying sub-requirement information c. The final target AI models selected are AI model A, AI model B, and AI model C, respectively.

[0087] Furthermore, to effectively improve the accuracy of selecting target artificial intelligence models, step S20 includes: acquiring functional description information of multiple artificial intelligence models accessed by the browser; determining the functions required to satisfy each sub-requirement information; matching the required functions with the functional description information; and selecting target artificial intelligence models from the multiple artificial intelligence models that satisfy each sub-requirement information based on the matching results.

[0088] It should be understood that functional description information refers to information describing the functions of various artificial intelligence models. For example, the function of artificial intelligence model A is to search for materials, and the function of artificial intelligence model B is to construct a framework. At this time, other intelligent models integrated in the browser will also be used to analyze and predict the functions required to meet each sub-requirement information. For example, the function required to meet sub-requirement b is the function of constructing a framework, and the function required to meet sub-requirement information c is the function of generating a PPT. Then, based on the matching results of the required functions and functional description information, the target artificial intelligence model that meets each sub-requirement information is selected from a variety of artificial intelligence models.

[0089] Furthermore, to ensure seamless and successful access to the browser for AI models from different sources and of different types, before step S20, the process includes: generating a standard multi-model access framework based on a unified interface protocol strategy and a unified data exchange strategy; and accessing the AI ​​model to be accessed to the browser based on the standard multi-model access framework.

[0090] It is understandable that the unified interface protocol strategy refers to a strategy for unifying interface protocols. Similarly, the unified data exchange strategy refers to a strategy for unifying data transformation. After generating the standard multi-model access framework, AI models that originally did not meet the access conditions can be connected to the browser based on the standard multi-model access framework. This ensures that AI models from different sources and of different types can be seamlessly and successfully connected to the browser. In addition, this standard multi-model access framework supports dynamic loading and unloading of models, enabling flexible combination and expansion of models.

[0091] Furthermore, to ensure seamless and successful access to the browser for AI models from different sources and of different types, the step of accessing the AI ​​model to be accessed to the browser based on the standard multi-model access framework includes: obtaining model feature information of each AI model supported by the browser; generating AI model access conditions based on the model feature information; obtaining attribute information of the AI ​​model to be accessed; and accessing the AI ​​model to be accessed to the browser based on the standard multi-model access framework, according to the attribute information and the AI ​​model access conditions.

[0092] It should be understood that model feature information refers to the feature information that identifies different models. This model feature information includes, but is not limited to, interface protocols, model formats, data types, etc. Artificial intelligence model access conditions refer to the conditions that enable seamless and successful access to the browser.

[0093] Furthermore, to ensure seamless and successful access to the browser for AI models from different sources and of different types, the step of accessing the AI ​​model to be accessed to the browser based on the standard multi-model access framework and according to the attribute information and the AI ​​model access conditions includes: determining whether there are any AI models among the AI ​​models to be accessed that do not meet the AI ​​model access conditions based on the attribute information; if so, performing data conversion on the AI ​​model that does not meet the AI ​​model access conditions based on the standard multi-model access framework; and accessing the converted AI model to the browser.

[0094] Understandably, after obtaining the attribute information of the AI ​​model to be accessed, it is determined whether any AI models do not meet the access conditions. If so, data conversion is required based on the standard multi-model access framework to ensure that the converted AI model meets the access conditions. Then, the converted AI model is accessed through the browser. Otherwise, the AI ​​model is directly accessed through the browser. This ensures that AI models from different sources and of different types can be seamlessly and successfully accessed through the browser, overcoming the shortcomings of incompatible data types and interface protocols that prevent seamless and successful browser access.

[0095] In addition, to reduce the browser's workload, some AI models that are used less frequently and have similar models can be dynamically uninstalled.

[0096] Step S30: Determine the collaborative operation mode between the target artificial intelligence models, and generate recommended content corresponding to the demand information based on the collaborative operation mode by the target artificial intelligence models.

[0097] It should be understood that the collaborative operation mode refers to the mode of controlling multiple target artificial intelligence models to run simultaneously. Under this collaborative operation mode, the overall efficiency of multiple target artificial intelligence models is relatively high. For example, there is no dependency between artificial intelligence model A and artificial intelligence model B, there is a dependency between artificial intelligence model A and artificial intelligence model C, and there is a dependency between artificial intelligence model B and artificial intelligence model C. In this case, the collaborative operation mode first controls artificial intelligence model A and artificial intelligence model B to run simultaneously, and then controls artificial intelligence model C to run.

[0098] It should be noted that once recommended content corresponding to the user's needs is generated, the recommended content will be displayed to the user in a timely manner, and the user will be provided with online editing and browsing functions. At the same time, a download button will also be provided to the user so that the user can edit and browse the recommended content offline.

[0099] Furthermore, to effectively improve the accuracy of generated recommended content, the step of generating recommended content corresponding to the demand information through the target artificial intelligence model based on the collaborative operation mode includes: controlling the collaborative operation of each target artificial intelligence model based on the collaborative operation mode, and obtaining the sub-recommendation content output by each target artificial intelligence model; generating recommended content corresponding to the demand information based on each sub-recommendation content.

[0100] Understandably, sub-recommended content refers to the recommended content output by each target AI model through analysis, searching, and construction. For example, AI model A outputs sub-recommended content as materials, and AI model B outputs sub-recommended content as a framework. After obtaining the various sub-recommended contents, the PPT creation model performs data fusion to generate recommended content corresponding to the demand information, such as a complete PPT.

[0101] It should be understood that during the operation of various artificial intelligence models based on a browser, this embodiment will also enable resource monitoring and early warning functions to provide timely alerts when resources are scarce or wasted. Simultaneously, regular security checks and vulnerability patching will be performed to ensure system security.

[0102] It should be noted that after showing recommended content to users, the system also collects user feedback data in real time, extracts medium and low feedback data from the feedback data, and then continuously optimizes the algorithm and parameter configuration based on the medium and low feedback data, regularly updates the artificial intelligence model and application, and improves intelligent analysis and decision-making capabilities.

[0103] This embodiment responds to a request triggered in a browser by breaking down the request information into sub-request information; selects target AI models from various AI models accessed by the browser that satisfy each sub-request information; determines the collaborative operation mode between the target AI models; and generates recommended content corresponding to the request information based on the collaborative operation mode. By breaking down the triggered request information into sub-request information, selecting target AI models that satisfy each sub-request information, and then using these target AI models to generate recommended content based on the collaborative operation mode, the accuracy of the generated recommended content can be effectively improved, thereby enhancing the user experience and meeting the complex and diverse needs of users.

[0104] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in the first embodiment described above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 2 Step S30 includes steps S301 to S303:

[0105] Step S301: Group the target artificial intelligence models to obtain a combination of target artificial intelligence models.

[0106] It should be understood that a target AI model combination refers to a combination of multiple target AI models. Taking two target AI models as an example, the target AI models are grouped into pairs, such as AI model A and AI model B as one group, AI model A and AI model C as one group, and AI model B and AI model C as one group.

[0107] Step S302: Dependency detection is performed on each of the target artificial intelligence model combinations.

[0108] Understandably, dependency detection refers to detecting whether there are dependencies between target AI models. For example, if the input parameters of AI model C are the output parameters of AI models A and B, it indicates that AI model C is dependent on AI models A and B respectively. The input parameters of AI model A are not affected by the output parameters of AI models B and C. Similarly, the input parameters of AI model B are not affected by the output parameters of AI models A and C, indicating that there is no dependency between AI models A and B.

[0109] Step S303: Determine the collaborative operation mode between the target artificial intelligence models based on the dependency detection results.

[0110] It should be understood that the collaborative operation mode refers to the mode of controlling multiple target artificial intelligence models to run simultaneously in a protocol. Under this collaborative operation mode, the overall efficiency of multiple target artificial intelligence models is relatively high.

[0111] Furthermore, to effectively improve the accuracy of determining the collaborative operation mode, step S303 includes: obtaining operational description information of each of the target artificial intelligence models; obtaining resource information required for the normal operation of each of the target artificial intelligence models based on the operational description information; and determining the collaborative operation mode between the target artificial intelligence models based on the dependency detection results and the required resource information.

[0112] It is understood that the runtime description information refers to the description information of running each target artificial intelligence model. This runtime description information includes, but is not limited to, the input parameters, output parameters, required resource information, accuracy, and training loss for normal operation. After obtaining the resource information required for the normal operation of each target artificial intelligence model, the collaborative operation mode between the target artificial intelligence models is determined by combining the dependency detection results.

[0113] Furthermore, to effectively improve the accuracy of determining the collaborative operation mode, the step of determining the collaborative operation mode between the target artificial intelligence models based on the dependency detection results and the required resource information includes: obtaining the remaining system resource information; determining the collaborative operation time of each target artificial intelligence model based on the dependency detection results, the required resource information, and the remaining system resource information; and determining the collaborative operation mode between the target artificial intelligence models based on the collaborative operation time.

[0114] It should be understood that the collaborative operation moment refers to the moment when the various target artificial intelligence models are controlled to operate collaboratively. That is, when the collaborative operation moment is reached, the operation of each target artificial intelligence model is controlled to begin. In order to avoid browser malfunction due to insufficient resources, it is necessary to determine the collaborative operation moment of each target artificial intelligence model in combination with the system's remaining resource information. For example, the collaborative operation moment of artificial intelligence model A and artificial intelligence model B is t1, and the collaborative operation moment of artificial intelligence model C is t2.

[0115] Furthermore, to effectively improve the accuracy of determining the collaborative operation mode, the step of determining the collaborative operation time of each target AI model based on the dependency detection results, the required resource information, and the remaining system resource information includes: obtaining target AI models with dependent relationships and target AI models without dependent relationships based on the dependency detection results; selecting target AI models that run synchronously with the target AI models without dependent relationships from among the target AI models with dependent relationships; determining the total resource information required by the synchronously running target AI models based on the required resource information; and inputting the total resource information, the remaining system resource information, and the required resource information into the collaborative operation time prediction model to obtain the collaborative operation time of each target AI model output by the collaborative operation time prediction model.

[0116] Understandably, after dependency detection, each target AI model is divided into target AI models with dependency relationships and target AI models without dependency relationships based on the dependency detection results. For example, there is no dependency relationship between AI model A and AI model B, there is a dependency relationship between AI model A and AI model C, and there is a dependency relationship between AI model B and AI model C.

[0117] It should be understood that, based on the demand information analysis, AI Model A and AI Model B are AI models running synchronously. At this point, it is necessary to further determine the total resource information required by the target AI models running synchronously based on the resource information required by each model. The total resource information, the remaining system resource information, and the resource information required by each model are then sequentially input into the collaborative operation time prediction model. The collaborative operation time prediction model predicts the collaborative operation time of each target AI model by analyzing the resource information and the collaborative process of multiple AI models. This collaborative operation time prediction model also uses an intelligent scheduling algorithm during training to make reasonable scheduling under limited resources, ensuring the collaborative operation of multiple AI models and maximizing the overall model's operating efficiency.

[0118] This embodiment obtains target AI model combinations by grouping the target AI models; performs dependency detection on each target AI model combination; and determines the collaborative operation mode between the target AI models based on the dependency detection results. By dividing the target AI models into target AI model combinations and obtaining the dependencies between models in each target AI model combination through dependency detection, and then considering the resource information required by each target AI model, the collaborative operation mode between the target AI models is determined. This effectively improves the accuracy of determining the collaborative operation mode and thus improves the model operation efficiency.

[0119] This application also provides a recommendation content generation apparatus, please refer to... Figure 4 The recommended content generation device includes:

[0120] The splitting module 10 is used to split the demand information triggered in the browser to obtain sub-demand information in response to the demand information.

[0121] The selection module 20 is used to select a target artificial intelligence model from a variety of artificial intelligence models accessed by the browser that respectively meet the sub-requirement information.

[0122] The generation module 30 is used to determine the collaborative operation mode between the target artificial intelligence models, and generate recommended content corresponding to the demand information based on the collaborative operation mode by the target artificial intelligence models.

[0123] This embodiment responds to a request triggered in a browser by breaking down the request information into sub-request information; selects target AI models from various AI models accessed by the browser that satisfy each sub-request information; determines the collaborative operation mode between the target AI models; and generates recommended content corresponding to the request information based on the collaborative operation mode. By breaking down the triggered request information into sub-request information, selecting target AI models that satisfy each sub-request information, and then using these target AI models to generate recommended content based on the collaborative operation mode, the accuracy of the generated recommended content can be effectively improved, thereby enhancing the user experience and meeting the complex and diverse needs of users.

[0124] The recommendation content generation apparatus provided in this application, employing the recommendation content generation method in the above embodiments, can solve the technical problem of low accuracy of recommendation content in the prior art. Compared with the prior art, the beneficial effects of the recommendation content generation apparatus provided in this application are the same as those of the recommendation content generation method provided in the above embodiments, and other technical features in the recommendation content generation apparatus are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0125] In one embodiment, the splitting module 10 is further configured to extract features from the demand information to obtain demand features; when the demand features are application display features, search for applications integrated on the browser; generate an application list based on each application and the attribute information of the application, and display the application list.

[0126] In one embodiment, the splitting module 10 is further configured to respond to application update information triggered in the browser, determine an application update data packet based on the application update information; extract the header of the application update data packet; determine the application to be updated in the application list based on the header; and update the application to be updated based on the application update data packet.

[0127] In one embodiment, the selection module 20 is further configured to acquire functional description information of multiple artificial intelligence models accessing the browser; determine the functions required to satisfy each sub-requirement information; match the required functions with the functional description information; and select a target artificial intelligence model from the multiple artificial intelligence models that satisfies each sub-requirement information based on the matching results.

[0128] In one embodiment, the selection module 20 is further configured to generate a standard multi-model access framework based on the interface protocol unified strategy and the data exchange unified strategy; and to access the artificial intelligence model to be accessed to the browser based on the standard multi-model access framework.

[0129] In one embodiment, the selection module 20 is further configured to obtain model feature information of each artificial intelligence model supported by the browser; generate artificial intelligence model access conditions based on the model feature information; obtain attribute information of the artificial intelligence model to be accessed; and, based on the standard multi-model access framework, access the artificial intelligence model to be accessed to the browser according to the attribute information and the artificial intelligence model access conditions.

[0130] In one embodiment, the selection module 20 is further configured to determine, based on the attribute information, whether there is an artificial intelligence model among the artificial intelligence models to be accessed that does not meet the access conditions for artificial intelligence models; if so, to perform data conversion on the artificial intelligence model that does not meet the access conditions for artificial intelligence models based on the standard multi-model access framework; and to access the converted artificial intelligence model to the browser.

[0131] In one embodiment, the generation module 30 is further configured to group the target artificial intelligence models to obtain target artificial intelligence model combinations; perform dependency detection on each target artificial intelligence model combination; and determine the collaborative operation mode between the target artificial intelligence models based on the dependency detection results.

[0132] In one embodiment, the generation module 30 is further configured to obtain runtime description information of each of the target artificial intelligence models; obtain resource information required for the normal operation of each of the target artificial intelligence models based on the runtime description information; and determine the collaborative operation mode between the target artificial intelligence models based on the dependency detection results and the required resource information.

[0133] In one embodiment, the generation module 30 is further configured to obtain system remaining resource information; determine the collaborative running time of each target artificial intelligence model based on the dependency detection result, the required resource information, and the system remaining resource information; and determine the collaborative running mode between the target artificial intelligence models based on the collaborative running time.

[0134] In one embodiment, the generation module 30 is further configured to: obtain target AI models with dependent relationships and target AI models without dependent relationships based on dependency detection results; select target AI models that run synchronously with the target AI models without dependent relationships from the target AI models with dependent relationships; determine the total resource information required by the synchronously running target AI models based on the required resource information for each of the specified resources; input the total resource information, the remaining system resource information, and the required resource information for each of the specified resources into the collaborative running time prediction model, and obtain the collaborative running time of each target AI model output by the collaborative running time prediction model.

[0135] In one embodiment, the generation module 30 is further configured to control the collaborative operation of each of the target artificial intelligence models based on the collaborative operation mode, and obtain the sub-recommendation content output by each of the target artificial intelligence models; and generate recommendation content corresponding to the demand information based on each of the sub-recommendation content.

[0136] This application provides a recommended content generation device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable 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 recommended content generation method in Embodiment 1 above.

[0137] The following is for reference. Figure 4 The diagram illustrates a structural schematic suitable for implementing the recommended content generation device of the embodiments of this application. The recommended content generation device in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 4 The recommended content generation device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0138] like Figure 4As shown, the recommended content generation device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the recommended content generation device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows the content generation device to communicate wirelessly or wiredly with other devices to exchange data. While the figures show content generation devices with various systems, it should be understood that implementing or having all of the systems shown is not required. More or fewer systems may be implemented alternatively.

[0139] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a 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, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0140] The recommendation content generation device provided in this application, employing the recommendation content generation method in the above embodiments, can solve the technical problem of low accuracy of recommendation content in the prior art. Compared with the prior art, the beneficial effects of the recommendation content generation device provided in this application are the same as those of the recommendation content generation method provided in the above embodiments, and other technical features in this recommendation content generation device are the same as those disclosed in the method of the previous embodiment, and will not be repeated here.

[0141] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0142] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0143] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the recommended content generation method in the above embodiments.

[0144] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having 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 thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0145] The aforementioned computer-readable storage medium may be included in the recommendation content generation device; or it may exist independently and not assembled into the recommendation content generation device.

[0146] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and 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).

[0147] 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 application. 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.

[0148] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0149] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described method for generating recommended content, thereby solving the technical problem of low accuracy of recommended content in the prior art. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the recommended content generation method provided in the above embodiments, and will not be repeated here.

[0150] This invention discloses A1. A method for generating recommended content, the method comprising:

[0151] In response to the request information triggered in the browser, the request information is split into sub-request information;

[0152] Select a target artificial intelligence model from the various artificial intelligence models accessed by the browser that respectively meet the sub-requirement information;

[0153] Determine the collaborative operation mode among the target artificial intelligence models, and generate recommended content corresponding to the demand information based on the collaborative operation mode using the target artificial intelligence models.

[0154] A2. As described in A1, the step of selecting a target artificial intelligence model from multiple artificial intelligence models accessing the browser that respectively meet the sub-requirement information includes:

[0155] Obtain functional description information of various artificial intelligence models accessed by the browser;

[0156] Determine the functions required to satisfy each of the sub-requirements;

[0157] The required functions are matched with the function description information respectively;

[0158] Based on the matching results, a target artificial intelligence model that satisfies each sub-requirement information is selected from the various artificial intelligence models.

[0159] A3. As described in A1, the step of determining the collaborative operation mode between the target artificial intelligence models includes:

[0160] The target artificial intelligence models are grouped to obtain a combination of target artificial intelligence models;

[0161] Dependency detection is performed on each of the aforementioned target AI model combinations;

[0162] The collaborative operation mode between the target artificial intelligence models is determined based on the dependency detection results.

[0163] A4. The method as described in A3, wherein the step of determining the collaborative operation mode between the target artificial intelligence models based on the dependency detection results includes:

[0164] Obtain the operational description information of each of the target artificial intelligence models;

[0165] Based on the operational description information, obtain the resource information required for the normal operation of each of the target artificial intelligence models;

[0166] The collaborative operation mode among the target artificial intelligence models is determined based on the dependency detection results and the required resource information.

[0167] A5. The method as described in A4, wherein the step of determining the collaborative operation mode among the target artificial intelligence models based on the dependency detection results and the required resource information includes:

[0168] Obtain information on remaining system resources;

[0169] The collaborative running time of each target artificial intelligence model is determined based on the dependency detection results, the required resource information, and the remaining system resource information.

[0170] The collaborative operation mode among the target artificial intelligence models is determined based on each of the collaborative operation moments.

[0171] A6. The method as described in A5, wherein the step of determining the collaborative running time of each target artificial intelligence model based on the dependency detection result, the required resource information, and the remaining system resource information includes:

[0172] Based on the dependency detection results, target AI models with and without dependency relationships are obtained;

[0173] Select a target AI model that runs synchronously with the target AI model that does not have a dependency relationship from among the target AI models that have a dependency relationship;

[0174] Determine the total resource information required for the target artificial intelligence model to run synchronously based on the resource information described above.

[0175] The total resource information, the remaining system resource information, and each of the required resource information are input into the collaborative operation time prediction model to obtain the collaborative operation time of each target artificial intelligence model output by the collaborative operation time prediction model.

[0176] A7. The method as described in A1, prior to the step of selecting a target artificial intelligence model from multiple artificial intelligence models accessing the browser that respectively meet the sub-requirement information, further includes:

[0177] A standard multi-model access framework is generated based on the unified strategy for interface protocols and the unified strategy for data exchange.

[0178] The AI ​​model to be accessed is connected to the browser based on the standard multi-model access framework.

[0179] A8. As described in A7, the step of connecting the AI ​​model to be accessed to the browser based on the standard multi-model access framework includes:

[0180] Obtain model feature information for each artificial intelligence model supported by the browser;

[0181] Generate AI model access conditions based on the model feature information;

[0182] Obtain the attribute information of the artificial intelligence model to be connected;

[0183] Based on the standard multi-model access framework, the artificial intelligence model to be accessed is accessed to the browser according to the attribute information and the access conditions of the artificial intelligence model.

[0184] A9. The method described in A8, wherein the step of connecting the artificial intelligence model to be connected to the browser based on the standard multi-model access framework and according to the attribute information and the artificial intelligence model access conditions includes:

[0185] Based on the attribute information, determine whether there are any artificial intelligence models among the artificial intelligence models to be accessed that do not meet the access conditions for artificial intelligence models;

[0186] If so, then data conversion is performed on the artificial intelligence models that do not meet the artificial intelligence model access conditions based on the standard multi-model access framework;

[0187] The converted AI model is then connected to the browser.

[0188] A10. The method as described in any one of A1 to A9, wherein the step of generating recommended content corresponding to the demand information based on the collaborative operation mode using the target artificial intelligence model includes:

[0189] Based on the collaborative operation mode, control the collaborative operation of each target artificial intelligence model and obtain the sub-recommendation content output by each target artificial intelligence model;

[0190] Based on each of the sub-recommendation contents, generate recommended content corresponding to the demand information.

[0191] A11. The method as described in A1, further comprising, after the step of responding to the request information triggered in the browser:

[0192] The demand information is subjected to feature extraction to obtain demand features;

[0193] When the required feature is an application display feature, search for applications integrated into the browser;

[0194] An application list is generated based on each application and its attribute information, and the application list is then displayed.

[0195] A12. As described in A11, after the step of displaying the application list, the method further includes:

[0196] In response to application update information triggered in the browser, an application update data packet is determined based on the application update information;

[0197] Extract the header of the application update data packet;

[0198] The application to be updated in the application list is determined based on the header;

[0199] The application to be updated is updated according to the application update data package.

[0200] The present invention also discloses B13. A recommended content generation apparatus, the apparatus comprising:

[0201] The splitting module is used to split the demand information triggered in the browser to obtain sub-demand information in response to the demand information.

[0202] The selection module is used to select a target artificial intelligence model that meets each sub-requirement information from a variety of artificial intelligence models accessed by the browser;

[0203] The generation module is used to determine the collaborative operation mode between the target artificial intelligence models, and generate recommended content corresponding to the demand information based on the collaborative operation mode by the target artificial intelligence models.

[0204] B14. The apparatus as described in B13, wherein the selection module is further configured to acquire functional description information of multiple artificial intelligence models accessing the browser; determine the functions required to satisfy each sub-requirement information; match the required functions with the functional description information; and select a target artificial intelligence model from the multiple artificial intelligence models that satisfies each sub-requirement information based on the matching results.

[0205] B15. The apparatus as described in B13, wherein the generation module is further configured to group the various target artificial intelligence models to obtain a combination of target artificial intelligence models; perform dependency detection on each combination of target artificial intelligence models respectively; and determine the collaborative operation mode between the target artificial intelligence models based on the dependency detection results.

[0206] B16. The apparatus as described in B15, wherein the generation module is further configured to acquire operational description information of each of the target artificial intelligence models; acquire resource information required for the normal operation of each of the target artificial intelligence models based on the operational description information; and determine a collaborative operation mode among the target artificial intelligence models based on the dependency detection results and the required resource information.

[0207] B17. The apparatus as described in B16, wherein the generation module is further configured to acquire system remaining resource information; determine the collaborative running time of each target artificial intelligence model based on the dependency detection result, the required resource information, and the system remaining resource information; and determine the collaborative running mode between the target artificial intelligence models based on the collaborative running time.

[0208] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the recommended content generation method described above.

[0209] The computer program product provided in this application can solve the technical problem of low accuracy of recommended content in existing technologies. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the recommended content generation method provided in the above embodiments, and will not be repeated here.

[0210] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for generating recommended content, characterized in that, The method includes: In response to the request information triggered in the browser, the request information is split into sub-request information; Select a target artificial intelligence model from the various artificial intelligence models accessed by the browser that respectively meet the sub-requirement information; Determine the collaborative operation mode among the target artificial intelligence models, and generate recommended content corresponding to the demand information based on the collaborative operation mode using the target artificial intelligence models.

2. The method as described in claim 1, characterized in that, The step of selecting a target artificial intelligence model from multiple artificial intelligence models accessed by the browser that respectively meets the sub-requirement information includes: Obtain functional description information of various artificial intelligence models accessed by the browser; Determine the functions required to satisfy each of the sub-requirements; The required functions are matched with the function description information respectively; Based on the matching results, a target artificial intelligence model that satisfies each sub-requirement information is selected from the various artificial intelligence models.

3. The method as described in claim 1, characterized in that, The step of determining the collaborative operation mode between the target artificial intelligence models includes: The target artificial intelligence models are grouped to obtain a combination of target artificial intelligence models; Dependency detection is performed on each of the aforementioned target AI model combinations; The collaborative operation mode between the target artificial intelligence models is determined based on the dependency detection results.

4. The method as described in claim 3, characterized in that, The step of determining the collaborative operation mode between the target artificial intelligence models based on the dependency detection results includes: Obtain the operational description information of each of the target artificial intelligence models; Based on the operational description information, obtain the resource information required for the normal operation of each of the target artificial intelligence models; The collaborative operation mode among the target artificial intelligence models is determined based on the dependency detection results and the required resource information.

5. The method as described in claim 4, characterized in that, The step of determining the collaborative operation mode among the target artificial intelligence models based on the dependency detection results and the required resource information includes: Obtain information on remaining system resources; The collaborative running time of each target artificial intelligence model is determined based on the dependency detection results, the required resource information, and the remaining system resource information. The collaborative operation mode among the target artificial intelligence models is determined based on each of the collaborative operation moments.

6. The method as described in claim 5, characterized in that, The step of determining the collaborative running time of each target artificial intelligence model based on the dependency detection results, the required resource information, and the remaining system resource information includes: Based on the dependency detection results, target AI models with and without dependency relationships are obtained; Select a target AI model that runs synchronously with the target AI model that does not have a dependency relationship from among the target AI models that have a dependency relationship; Determine the total resource information required for the target artificial intelligence model to run synchronously based on the resource information described above. The total resource information, the remaining system resource information, and each of the required resource information are input into the collaborative operation time prediction model to obtain the collaborative operation time of each target artificial intelligence model output by the collaborative operation time prediction model.

7. A recommended content generation device, characterized in that, The device includes: The splitting module is used to split the demand information triggered in the browser to obtain sub-demand information in response to the demand information. The selection module is used to select a target artificial intelligence model that meets each sub-requirement information from a variety of artificial intelligence models accessed by the browser; The generation module is used to determine the collaborative operation mode between the target artificial intelligence models, and generate recommended content corresponding to the demand information based on the collaborative operation mode by the target artificial intelligence models.

8. A recommendation content generation device, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the recommendation content generation method as described in any one of claims 1 to 6.

9. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the recommendation content generation method as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the recommendation content generation method as described in any one of claims 1 to 6.

Images