A resource publishing method and device based on LLM model

Abstract

A resource publishing method based on an LLM model is applied to a resource publishing platform. The resource publishing platform executes a conversion task of converting an external resource into a standard resource supported by the resource publishing platform based on an accessed LLM model. The method comprises: obtaining resource information of the external resource to be published; generating a prompt word corresponding to the conversion task based on the resource information of the external resource, and inputting the prompt word into an LLM model so that the LLM model executes an inference calculation corresponding to the conversion task based on the prompt word to convert the external resource into the standard resource; obtaining the standard resource output by the LLM model, and publishing the standard resource on the resource publishing platform.

Description

Technical Field

[0001] The embodiments of this specification relate to the field of artificial intelligence technology, and particularly to a resource publishing method and device based on the LLM model. Background Art

[0002] On some traditional resource publishing platforms, in order to convert external resources into standard resources supported by the resource publishing platform and publish them on the resource publishing platform, a large amount of manual operations are usually required.

[0003] For example, usually, the operation staff responsible for resource publishing needs to obtain the resource information of the external resources offline from the provider of the external resources, create a data structure of a standard resource corresponding to the external resources on the resource publishing platform, and then fill in the detailed content of the obtained resource information of the external resources into the created data structure of the standard resource one by one. The whole process is time-consuming and laborious. Summary of the Invention

[0004] This specification proposes a resource publishing method based on the LLM model, which is applied to a resource publishing platform. The resource publishing platform performs a conversion task of converting external resources into standard resources supported by the resource publishing platform based on the accessed LLM model. The method includes:

[0005] Obtain the resource information of the external resources to be published;

[0006] Based on the resource information of the external resources, generate a prompt corresponding to the conversion task, and input the prompt into the LLM model, so that the LLM model performs inference calculations corresponding to the conversion task based on the prompt, and converts the external resources into the standard resources;

[0007] Obtain the standard resources output by the LLM model, and publish the standard resources on the resource publishing platform.

[0008] This specification also proposes a resource publishing device based on the LLM model, which is applied to a resource publishing platform. The resource publishing platform performs a conversion task of converting external resources into standard resources supported by the resource publishing platform based on the accessed LLM model. The method includes:

[0009] An obtaining module, which obtains the resource information of the external resources to be published;

[0010] A conversion module, which generates a prompt corresponding to the conversion task based on the resource information of the external resources, and inputs the prompt into the LLM model, so that the LLM model performs inference calculations corresponding to the conversion task based on the prompt, and converts the external resources into the standard resources;

[0011] In the above embodiments, by introducing an LLM model in the resource publishing platform to perform the conversion task of converting external resources into standard resources supported by the resource publishing platform, when publishing external resources, prompt words can be generated to guide the LLM model to convert external resources into standard resources, so that external resources can be automatically converted into standard resources for publishing without manual participation, thereby reducing labor costs and improving the publishing efficiency of external resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions of the embodiments of this specification, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments recorded in this specification. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0013] Figure 1 It is a schematic diagram of the architecture of an Internet service system shown in an embodiment of this specification;

[0014] Figure 2 It is a flowchart of a resource publishing method based on an LLM model shown in an embodiment of this specification;

[0015] Figure 3 It is a schematic diagram of an event-driven task execution framework shown in an embodiment of this specification;

[0016] Figure 4 It is a flowchart of converting external resources into standard resources shown in an embodiment of this specification;

[0017] Figure 5 It is an architecture diagram of a task execution framework based on a general LLM model shown in an embodiment of this specification;

[0018] Figure 6 It is a schematic structural diagram of an electronic device shown in an embodiment of this specification;

[0019] Figure 7 It is a block diagram of a resource publishing device based on an LLM model shown in an embodiment of this specification. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] Exemplary embodiments will be described in detail herein, and examples thereof are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with one or more embodiments of this specification. On the contrary, they are merely examples of devices and methods consistent with some aspects of one or more embodiments of this specification as detailed in the appended claims.

[0021] It should be noted that: In other embodiments, the steps of the corresponding methods are not necessarily executed in the order shown and described in this specification. In some other embodiments, the steps included in the method may be more or fewer than those described in this specification. In addition, a single step described in this specification may be decomposed into multiple steps for description in other embodiments; and multiple steps described in this specification may also be combined into a single step for description in other embodiments.

[0022] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this specification are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of the relevant countries and regions, and corresponding operation entrances are provided for users to choose to authorize or refuse.

[0023] This specification aims to propose a technical solution that, without manual intervention, automatically converts high-value external resources into standard resources for publication on a resource publishing platform by introducing an LLM (Large Language Model) model into the resource publishing platform.

[0024] Figure 1 is a schematic diagram of the architecture of an Internet service system provided by an exemplary embodiment. As Figure 1 shown, the system may include a server 11, a network 12, and several electronic devices, such as a PC (Personal Computer) 13, a mobile phone 14, etc.

[0025] The server 11 may be a physical server including an independent host, or the server 11 may be a virtual server hosted by a host cluster. During operation, the server 11 may run the server-side program of an application to implement the relevant functions of the application. For example, when the server 11 runs the program of an Internet service, it may be implemented as the corresponding Internet service platform.

[0026] PC 13 and mobile phone 14 are only some types of electronic devices that users can use. In fact, users can obviously also use electronic devices of the following types: tablet devices, laptop computers, personal digital assistants (PDAs), wearable devices (such as smart glasses, smart watches, etc.). One or more embodiments of this specification do not limit this. During operation, the electronic device can run the program on the client side of a certain application to implement the related functions of the application. For example, when the electronic device runs the program of an Internet service, it can be implemented as the client of the Internet service. Among them, the application program of the client side of the above Internet service can be started and run on the electronic device. The program on the client side can be a native application installed on the electronic device, or the program on the client side can be in the form of a small program, a fast application, or other similar forms.

[0027] Of course, when using web technologies such as HTML5 or similar, relevant functions can be implemented through the pages displayed by the browser. Here, the browser can be an independent browser application or a browser module embedded in some applications.

[0028] For the network 12 for the interaction between electronic devices such as PC 13 and mobile phone 14 and the server 11, it can be specifically selected to use wired or wireless networks to achieve communication based on the communication methods supported by the corresponding electronic devices. This specification does not limit this. For example, PC 13 can support both wired and wireless communications, so wired or wireless networks can be used to achieve communication according to needs, while mobile phone 14 usually only supports wireless communication, so wireless networks can be used to achieve communication.

[0029] It should be noted that: the above Internet service can include any service implemented on the Internet; for example, the above Internet service can specifically include resource publishing services (such as commodity publishing platforms) that can provide resources (such as travel-related commodities) for users, resource trading services (such as travel trading services), and other e-commerce services, or other types of services. This specification does not limit this specifically.

[0030] The technical solutions of this specification will be described in detail below with reference to the accompanying drawings.

[0031] Please refer to Figure 2 , Figure 2 which is a flowchart of a resource publishing method based on the LLM model shown in this specification. The method includes the following execution processes:

[0032] Step 202: Obtain the resource information of the external resource to be published;

[0033] The execution subject of the above resource publishing method can specifically be a resource publishing platform. Among them, in practical applications, this resource publishing platform can specifically include a service platform capable of publishing any type of resource.

[0034] In an illustrated embodiment, the above resources can specifically include commodity - type resources. In this case, the above resource publishing platform can specifically include a commodity publishing platform.

[0035] For example, in one example, the above commodity can be a travel - related commodity, and the above commodity publishing platform can be a travel transaction service platform that can provide travel transaction services to users. It should be noted that the above travel - related commodities can specifically include any type of travel - related commodities; for example, ticket - type commodities, air - ticket - type commodities, accommodation - type commodities, train - ticket - type commodities, and so on.

[0036] Of course, in practical applications, the above resource publishing platform can also specifically be other types of resource publishing platforms other than the commodity publishing platform, which will not be listed one by one in this specification. For example, the above resource publishing platform can also be a content publishing platform for media content such as short videos, news, or other forms of content.

[0037] In this specification, in order to improve the automation level of the resource publishing platform when publishing external resources, this resource publishing platform can specifically access the LLM model and perform the conversion task of converting external resources into standard resources supported by the resource publishing platform through this LLM model.

[0038] Among them, the above LLM model can specifically include a general LLM model, or an LLM model obtained by fine - tuning and training on the basis of a general LLM model with a dataset in a specific application scenario, which is not specifically limited in this specification.

[0039] The above external resources can specifically include resources provided by an external resource provider docked with the above resource publishing platform; correspondingly, the above standard resources can specifically include resources with a standard information format supported by the resource publishing platform.

[0040] For example, in one example, taking the above resources as commodity - type resources, the above external resources can specifically be commodities provided by an external commodity provider (such as an external merchant cooperating with the commodity publishing platform) docked with the commodity publishing platform. And the above standard resources can specifically be commodities with a standard information format supported by the commodity publishing platform.

[0041] When the resource publishing platform supports publishing external resources provided by an external resource provider, it can perform system docking with the external resource provider to obtain the resource information of the external resources to be published from the external resource provider.

[0042] For example, in one example, still taking the above-mentioned resource as a resource of the commodity category as an example, at this time, the above-mentioned resource publishing platform can specifically be a commodity publishing platform, and the above-mentioned external resource provider can be an external merchant cooperating with the commodity publishing platform. In this case, an API interface for obtaining external commodities provided by the external merchant can be provided on the server side of the external merchant. When the resource publishing platform needs to publish the external commodities provided by the external merchant, it can obtain the commodity information of the external commodities provided by the external merchant from the server side of the external merchant by calling this API interface.

[0043] It should be noted that the resource information of the external resources to be published obtained by the resource publishing platform from the external resource provider can specifically include any form of information related to resource conversion, which is not particularly limited in this specification.

[0044] For example, in one example, taking the above-mentioned resource as a ticket commodity provided by an external commodity provider as an example, the resource information of the above-mentioned external resources that needs to be obtained can specifically include information such as the name of the ticket product, the name of the scenic spot, the country where the scenic spot is located, the city where the scenic spot is located, the address of the scenic spot, and the cost of the scenic spot.

[0045] In practical applications, if the number of external resources that an external resource provider can provide is relatively large, in order to prevent the resource publishing platform that is system-connected to it from quickly obtaining all the external resources that the external resource provider can provide in a short time, the external resource provider usually performs flow limiting processing on the resource publishing platform that is system-connected to it.

[0046] For example, in one example, still taking the above-mentioned resource as a resource of the commodity category as an example, at this time, the above-mentioned resource publishing platform can specifically be a commodity publishing platform, and the above-mentioned external resource provider can be an external merchant cooperating with the commodity publishing platform. In this case, an API interface for obtaining external commodities provided by the external merchant can be provided on the server side of the external merchant, and flow limiting processing is performed on this API interface; for example, a maximum QPS (Queries Per Second) is configured for this API. When the resource publishing platform needs to publish the external commodities provided by the external merchant, it can obtain the commodity information of the external commodities by calling this API interface according to the maximum QPS configured for this API.

[0047] In this case, the above-mentioned resource publishing platform can specifically obtain as many high-value external resources provided by the external resource provider as possible on the premise that the external resource provider has performed flow limiting processing.

[0048] In an illustrated embodiment, when the above-mentioned resource publishing platform obtains the resource information of the external resources to be published from an external resource provider, it can specifically pre-calculate the resource value scores of each external resource in the external resource set provided by the external resource provider; wherein, the resource value score can specifically be used to represent the resource value of the external resource; then, the calculated resource value score can be further mapped into a scheduling priority; wherein, the scheduling priority is specifically in a positive correlation with the resource value score.

[0049] After mapping the calculated resource value score into a scheduling priority, the above-mentioned resource publishing platform can use the scheduling priority as a scheduling factor and adopt a periodic resource scheduling method to periodically perform resource scheduling on the external resources in the external resource set, so as to determine the external resources to be published from the external resource set, and then selectively obtain the resource information of these determined external resources to be published from the external resource provider.

[0050] Among them, the scoring strategy adopted by the above-mentioned resource publishing platform when calculating the resource value scores of each external resource in the external resource set provided by the external resource provider is not specifically limited in this specification. In actual applications, the resource value of each external resource in the external resource set can be flexibly evaluated in combination with specific requirements.

[0051] In an illustrated embodiment, the above-mentioned resource publishing platform can specifically evaluate the resource value of external resources by collecting resource indicators that can reflect the resource value of external resources.

[0052] In this case, the above-mentioned resource publishing platform can first obtain multiple resource indicators of each external resource in the external resource set; wherein, the multiple resource indicators can specifically be indicators that can represent the resource value of external resources; then, the multiple resource indicators can be weighted and calculated, and the weighted sum obtained from the weighted calculation can be used as the resource value score of each external resource in the external resource set.

[0053] In an illustrated embodiment, the multiple resource indicators can specifically include a first indicator for representing the traffic value of external resources and a second indicator for representing the transaction value of external resources; in this case, the above-mentioned resource publishing platform can perform weighted calculation on the first indicator and the second indicator to obtain the resource value score of each external resource in the external resource set.

[0054] Among them, the specific representation forms of the above-mentioned first index and the above-mentioned second index are not specifically limited in this specification. In practical applications, any index form that can be used to represent the traffic value of external resources can be used as the above-mentioned first index; correspondingly, any index form that can be used to represent the transaction value of external resources can also be used as the above-mentioned second index.

[0055] For example, in one example, the ratio of the access volume of an external resource to the total access volume of each external resource in the external resource set can be used to represent the traffic value of the external resource; in this case, the above-mentioned first index can specifically include the ratio of the access volume of an external resource to the total access volume of each external resource in the external resource set.

[0056] In another example, the product of the order volume of an external resource and the order amount (i.e., the order unit price) of the external resource can also be used to represent the transaction value of the external resource. In this case, the above-mentioned second index can specifically include the product of the order volume of an external resource and the order amount of the external resource.

[0057] It should be noted that in practical applications, when performing weighted calculations on the above-mentioned first index and the above-mentioned second index, the weighting coefficients of the first index and the second index can be flexibly set based on specific requirements and are not specifically limited in this specification.

[0058] In an illustrated implementation manner, the weighting coefficients of the first index and the second index can specifically be the weighting coefficients calculated based on the resource conversion rate of the external resource.

[0059] For example, in one example, the ratio of the weighting coefficient of the first index to the weighting coefficient of the second index can specifically be the resource conversion rate of the external resource; that is, when setting the weighting coefficients for the first index and the second index, it can be ensured that the ratio of the first index to the second index is the resource conversion rate of the external resource; for example, assuming that the resource conversion rate of the external resource is 5%, then it is only necessary to ensure that the ratio of the set first index to the second index is 0.05.

[0060] In an illustrated implementation manner, in addition to the first index for representing the traffic value of external resources and the second index for representing the transaction value of external resources, the above-mentioned multiple resource indexes can also include a third index for representing the change frequency of the resource information of the external resource. In this case, the above-mentioned resource publishing platform can perform weighted calculations on the first index, the second index, and the third index to obtain the resource value scores of each external resource in the external resource set.

[0061] In this way, during the process of calculating the resource value score of external resources, an indicator representing the change frequency of the resource information of external resources can be introduced, so that higher resource value scores can be given to those external resources whose resource information changes frequently. When the above resource publishing platform performs resource scheduling for the external resources in the external resource set provided by the external resource provider, it can preferentially obtain the resource information of those external resources whose resource information will change frequently, thus avoiding the problem of order placement failure caused by the failure to timely obtain the resource information of external resources whose resource information changes frequently.

[0062] Among them, the specific representation form of the above third indicator is not particularly limited in this specification. In practical applications, any form of indicator that can represent the change frequency of the resource information of external resources can be used as the above third indicator.

[0063] For example, in one example, the number of times each information field included in the resource information of an external resource changes within a preset time period in the future can be used to represent the change frequency of the resource information of the external resource. In this case, the above third indicator can specifically include the weighted sum obtained by performing weighted calculation on the number of times each information field included in the resource information of the external resource changes within a preset time period in the future.

[0064] Among them, when performing weighted calculation on the number of times each information field included in the resource information of an external resource changes within a preset time period in the future, the weighted values of the number of times these information fields change within a preset time period in the future can be flexibly set based on specific requirements and are not specifically limited in this specification.

[0065] For example, the weighted coefficients can be set for each field based on the importance of each field; and the importance of each field can specifically be evaluated by the degree of influence of each field on the transaction success rate of external resources. For example, for a field with a relatively high degree of influence on the transaction success rate of external resources, a larger weighted coefficient can be set; conversely, a smaller weighted coefficient can be set.

[0066] It should be noted that in practical applications, the number of times each information field included in the resource information of the above external resource changes within a preset time period in the future can specifically be the number predicted based on the historical number of times each information field included in the resource information of the external resource changed in the past historical period.

[0067] For example, in one example, the historical number of times each information field included in the resource information of an external resource has changed in the past historical period can be collected to construct a time series model, and then based on this time series model, the number of times each information field included in the resource information of the external resource will change within a preset duration in the future can be predicted.

[0068] When calculating the weighted values of the above first index, the above second index, and the above third index, the weighting coefficient of the third index can also be flexibly set based on specific requirements and will not be specifically limited in this specification.

[0069] In an illustrated embodiment, the weighting coefficient of the third index can specifically be a weighting coefficient calculated based on the order information of orders for which transactions with respect to the external resource have failed due to changes in the resource information.

[0070] For example, in one example, the above order information can specifically be the total order amount of orders for which transactions with respect to the external resource have failed due to changes in the resource information, and the weighting coefficient of the above third index can be a weighting coefficient calculated based on the total order amount of orders for which transactions with respect to the external resource have failed due to changes in the resource information.

[0071] It should be noted that the specific calculation method for calculating the weighting coefficient of the third index based on the total order amount of orders for which transactions with respect to the external resource have failed due to changes in the resource information will not be specifically limited in this specification. In practical applications, the calculation method can be flexibly constructed based on specific requirements.

[0072] For example, in one example, the weighting coefficient of the third index can specifically be calculated using the following expression:

[0073] K =

[0074] In the above expression, K represents the weighting coefficient of the above third index; N represents the order quantity of the external resource; T represents the total order amount of orders for which transactions with respect to the external resource have failed due to changes in the resource information; and M represents the total order amount of orders for which transactions with respect to the external resource have been successful.

[0075] It should also be noted that the mapping method adopted by the above resource publishing platform when further mapping the calculated resource value scores of each external resource into scheduling priorities will not be specifically limited in this specification. In practical applications, a suitable mapping strategy can be flexibly selected in combination with specific requirements.

[0076] In an illustrated embodiment, the above-mentioned resource publishing platform may specifically compare the resource value score of an external resource with the maximum and minimum resource value scores of all external resources in the external resource set that the external resource provider can provide, so as to map the resource value score of the external resource into a scheduling priority.

[0077] For example, in an example, the above-mentioned resource publishing platform may use the following expression to further map the calculated resource value score of an external resource into a scheduling priority:

[0078] P =

[0079] In this expression, P represents the scheduling priority of the external resource; S represents the resource value score of the external resource; min represents the minimum resource value score of all external resources in the external resource set; max represents the maximum resource value score of all external resources in the above-mentioned external resource set; g represents the number of priority levels preset for the above-mentioned scheduling priority (i.e., the number of levels of priority). It should be noted that the number of priority levels preset for the above-mentioned scheduling priority can be flexibly set based on specific requirements in actual applications. For example, the number of levels of scheduling priority can be divided based on the total number of external resources in the above-mentioned external resource set.

[0080] Among them, the scheduling method adopted by the above-mentioned resource publishing platform for resource scheduling of the external resources in the external resource set periodically based on the scheduling priority of each external resource will not be specifically limited in this specification. In actual applications, a suitable scheduling strategy can be flexibly selected according to specific requirements.

[0081] In an illustrated embodiment, in order to improve the efficiency during resource scheduling, the above-mentioned resource publishing platform may organize the external resources in the external resource set into the form of a binary tree, and complete the resource scheduling for the external resources in the external resource set by traversing the external resources on the binary tree.

[0082] In this case, when the above-mentioned resource publishing platform performs resource scheduling for the external resources in the external resource set periodically based on the scheduling priority, it may pre-create scheduling nodes corresponding to each external resource in the external resource set; among them, the information recorded in the scheduling node may specifically include the scheduling priority of the external resource.

[0083] For example, in one example, the information recorded in the scheduling node may include, in addition to the scheduling priority of the external resource, in practical applications, the resource information of the external resource, the number of times the external resource has been scheduled, the time stamp corresponding to the moment of the last scheduling, and so on.

[0084] After creating scheduling nodes corresponding to each external resource in the external resource set, the above-mentioned scheduling priority included in the information recorded in each scheduling node can be used as a sorting key (key), and a binary tree can be constructed from the scheduling nodes corresponding to all the external resources in the above-mentioned external resource set.

[0085] After the binary tree is constructed, the above-mentioned resource publishing platform can periodically traverse the binary tree to find at least one target scheduling node with the highest recorded scheduling priority on the binary tree; for example, N scheduling nodes with the highest recorded scheduling priority on the binary tree can be found; the threshold N can be flexibly set based on specific requirements.

[0086] Then, the external resources corresponding to the above-mentioned at least one target scheduling node found can be determined as the external resources to be published, and the resource information of the external resources corresponding to the above-mentioned at least one target scheduling node found can be obtained from the external resource provider.

[0087] Among them, the specific type of the above-mentioned binary tree is not particularly limited in this specification; for example, in practical applications, a Red Black Tree can be used. Among them, a Red Black Tree is a self-balancing binary tree, and the specific process of constructing a Red Black Tree from the scheduling nodes corresponding to all the external resources in the above-mentioned external resource set will not be described in detail in this specification, and those skilled in the art can refer to the records in related technologies.

[0088] In one shown embodiment, in order to further improve the efficiency of resource scheduling, the scheduling priority of the external resources in the above-mentioned external resource set can be further mapped into a scheduling weight value, and the scheduling weight value can be used as a sorting key to construct a binary tree.

[0089] It should be noted that since the above-mentioned scheduling priority usually presets several different priority levels, in practical applications, different scheduling weight values can be set for different priority levels. At this time, when further mapping the scheduling priority of the external resources in the above-mentioned external resource set into a scheduling weight value, specifically, the priority level of the calculated scheduling priority of the external resources in the above-mentioned external resource set can be first determined, and then the scheduling weight value set for this priority level can be determined as the scheduling weight value corresponding to this scheduling priority.

[0090] In this case, the information recorded in the above scheduling node may further include a scheduling weight value corresponding to the priority level of the scheduling priority of each external resource; wherein, the scheduling priority of the external resource may specifically have a positive correlation with the scheduling weight value. When constructing the binary tree, the resource publishing platform may specifically use the scheduling weight value obtained by mapping the scheduling priority of the external resource as the sorting key, and construct a binary tree with the scheduling nodes corresponding to all the external resources in the above external resource set.

[0091] After the binary tree is constructed, the resource publishing platform may periodically traverse the binary tree to find at least one target scheduling node with the highest recorded scheduling weight value on the binary tree; then, the external resources corresponding to the at least one target scheduling node found may be determined as the external resources to be published, and the resource information of the external resources corresponding to the at least one target scheduling node found may be obtained from the external resource provider.

[0092] By further mapping the scheduling priority of the external resource into a scheduling weight value corresponding to the priority level of the scheduling priority, and performing resource scheduling based on the scheduling weight value, it is possible to no longer perform resource scheduling in units of external resources, but to batch-schedule external resources in units of priority levels, thereby further improving the efficiency of resource scheduling.

[0093] For example, since the above scheduling weight value usually corresponds to a class of external resources with a scheduling priority at a certain priority level, performing resource scheduling based on the above scheduling weight value is equivalent to batch-scheduling a class of external resources with a scheduling priority at that priority level.

[0094] In an illustrated embodiment, after the resource publishing platform obtains the resource information of the external resources corresponding to the at least one target scheduling node found from the external resource provider, in order to achieve balanced resource scheduling and avoid repeatedly scheduling the same external resource, the scheduling priority included in the information recorded in the at least one target scheduling node may further be updated based on a preset scheduling balance strategy.

[0095] On the one hand, if the sorting key of the above binary tree is the above scheduling priority, in this case, after the resource publishing platform obtains the resource information of the external resources corresponding to the at least one target scheduling node found from the external resource provider, it can update the scheduling priority recorded in the at least one target scheduling node based on a preset scheduling balance strategy. Then, it can re-determine the positions of the at least one target scheduling node in the binary tree based on the updated scheduling priority, and re-insert the at least one target scheduling node into the binary tree according to the re-determined positions, waiting for the execution of the scheduling in subsequent rounds.

[0096] On the other hand, if the sorting key of the above binary tree is the above scheduling weight value, in this case, after the resource publishing platform obtains the resource information of the external resources corresponding to the at least one target scheduling node found from the external resource provider, it can update the above scheduling priority recorded in the at least one target scheduling node based on a preset scheduling balance strategy, and synchronously update the above scheduling weight value recorded in the at least one target scheduling node based on the updated above scheduling priority; for example, in the same way as above, remap the updated above scheduling priority to a scheduling weight value. Then, it can re-determine the positions of the at least one target scheduling node in the binary tree based on the updated above scheduling weight value, and re-insert the at least one target scheduling node into the binary tree according to the re-determined positions, waiting for the execution of the scheduling in subsequent rounds.

[0097] Among them, the above scheduling balance strategy can specifically include any form of update strategy that can achieve balanced scheduling of external resources.

[0098] For example, it can make the scheduling priority of external resources increase or decrease uniformly in an increasing or decreasing direction. For instance, in an example, after the resource publishing platform obtains the resource information of the external resources corresponding to the at least one target scheduling node found from the external resource provider, it can uniformly increase the scheduling priority of the external resources corresponding to the at least one target scheduling node by the same value based on the original priority value of the scheduling priority.

[0099] In this way, since every time the resource publishing platform obtains the resource information of the external resources corresponding to the at least one target scheduling node found from the external resource provider, the scheduling priority of the external resources corresponding to the at least one target scheduling node will increase or decrease uniformly, it can schedule other external resources with the highest external priority or scheduling weight value in the next round of periodic scheduling, so that the same external resource can be scheduled repeatedly, and balanced scheduling of external resources can be achieved.

[0100] Step 104: Generate a prompt corresponding to the conversion task based on the resource information of the external resource, and input the prompt into the LLM model, so that the LLM model performs inference calculations corresponding to the conversion task based on the prompt, and convert the external resource into the standard resource;

[0101] After the resource publishing platform obtains the resource information of the external resource to be published, it can generate a prompt corresponding to the above conversion task based on the resource information of the external resource, and then input the prompt into the LLM model. After receiving the input prompt, the LLM model can perform inference calculations corresponding to the conversion task based on the prompt, and convert the external resource into the standard resource.

[0102] It should be noted that the content of the prompt included in the generated prompt is not specifically limited in this specification. In practical applications, it can be flexibly defined based on specific requirements.

[0103] In an illustrated embodiment, in order to address the problem that when the LLM model performs inference calculations on some complex tasks, a long task context may cause the inference ability of the LLM model to deteriorate. On the one hand, the execution link of the conversion task can be split by imitating the process of manually performing the conversion task, and the conversion task can be split into multiple subtasks. In this case, the conversion task can specifically include multiple subtasks executed in sequence. On the other hand, the execution of the multiple subtasks can also be driven by event triggering, and the multiple subtasks can be linked into a complete execution link. In this case, the multiple subtasks included in the conversion task can specifically be tasks executed based on event triggering.

[0104] In this case, when using the above LLM model to perform the above conversion task, first, a first prompt corresponding to the first subtask included in the conversion task can be generated based on the obtained resource information of the external resource to be published; where the first subtask can specifically refer to the first subtask among the multiple subtasks included in the conversion task. Then, the first prompt can be input into the LLM model, so that the LLM model performs inference calculations corresponding to the first subtask based on the first prompt;

[0105] After the first subtask is completed, it is possible to continue to obtain the first execution result corresponding to the first subtask output by the LLM model and generate a first event for triggering the execution of the second subtask included in the conversion task; wherein, the first subtask may specifically refer to the next subtask of the above-mentioned first subtask. Then, it is possible to determine whether the first event meets the triggering condition; if the first event meets the triggering condition, at this time, the first event can be executed, and further, a second prompt word corresponding to the second subtask can be generated based on the first execution result, and the second prompt word can be input into the LLM model so that the LLM model performs inference calculations corresponding to the second subtask based on the second prompt word.

[0106] After the second subtask is completed, at this time, the same process can be executed, continue to obtain the second execution result corresponding to the second subtask, and repeat the above process until all the subtasks included in the conversion task are completed, and then stop the above execution process when the external resource is successfully converted into a standard resource supported by the resource publishing platform.

[0107] In the above embodiments, since the execution link of the above conversion task is split into multiple subtasks, and the multiple subtasks are linked into a complete execution link by means of event triggering, the relatively long task context of the conversion task can be split into a form that is easier for the LLM model to understand and executed in slices. Therefore, in this way, on the one hand, the problem that the inference ability of the LLM model deteriorates for tasks with relatively long task contexts can be overcome, thereby improving the accuracy of the inference results of the LLM model; on the other hand, some general LLM models that have not been fine-tuned and do not have professional business knowledge can also be used to execute the above conversion task, thereby reducing the implementation cost of the above conversion task and enhancing the generality of the overall solution.

[0108] It should be noted that since the first subtask included in the above conversion task is the first subtask in the conversion task, the first subtask can usually be manually triggered and executed by the initiator of the conversion task, while the subtasks after the first subtask can be triggered and executed in the foregoing manner based on the event generated after the previous subtask is completed.

[0109] In practical applications, when the resource publishing platform drives the execution of the multiple subtasks by means of event triggering, it can specifically be implemented by using an event-driven task execution framework.

[0110] Please refer to Figure 3 , Figure 3 which is a schematic diagram of an event-driven task execution framework shown in this specification.

[0111] As Figure 3As shown, in this task execution framework, it can specifically include nodes such as event producers, event routers, event consumers, and event triggers.

[0112] Among them, the event producer is specifically used to generate events for triggering task execution.

[0113] The event router is used to pass events from the event producer to the appropriate consumers.

[0114] The event consumer is used to execute events that meet the trigger conditions.

[0115] The event trigger is used to detect whether the events generated by the event producer meet the trigger conditions.

[0116] For example, as Figure 3 shown, in practical applications, the event trigger can manage the events generated by the event producer in a publish-subscribe mode, and the event consumer can subscribe to the events generated by a specific event producer from the event trigger through a subscription mechanism. When the event trigger detects that the events generated by a certain event producer meet the trigger conditions, based on the distribution mechanism, the events that meet the trigger conditions can be accurately distributed to the event consumers that have subscribed to the events through the event router, and the event consumers will execute them.

[0117] It should be noted that in practical applications, the software forms of the above-mentioned event producer, event router, event consumer, and event trigger are not specifically limited in this specification.

[0118] For example, the above-mentioned event trigger can usually be an independent background service program on the service platform, running as an independent daemon process or microservice; the above-mentioned event producer can usually run in the form of an integrated component embedded in the business logic of the application program; the above-mentioned event router can usually run in the form of middleware; and the above-mentioned event consumer can usually run in the form of a microservice or an independent thread.

[0119] In this case, when the above-mentioned first subtask is completed, at this time, the above-mentioned event producer can obtain the first execution result corresponding to the first subtask output by the LLM model, and generate a first event for triggering the execution of the second subtask included in the conversion task based on the obtained first execution result; among them, the obtained first execution result can be included in the first event.

[0120] Then, the above event trigger can determine whether the first event meets the trigger condition; if the first event meets the trigger condition, at this time, the above event trigger can, based on the distribution mechanism, accurately distribute the event that meets the trigger condition to the event consumers subscribed to the event through event routing, and the event consumers will execute the first event. Further, a second prompt word corresponding to the second subtask is generated based on the above first execution result included in the first event, and the second prompt word is input into the LLM model so that the LLM model performs inference calculations corresponding to the second subtask based on the second prompt word.

[0121] After the second subtask is executed, the above event producer can continue to obtain the second execution result corresponding to the second subtask, and so on, repeating the above process until all the subtasks included in the conversion task are executed, and then stop the above execution process when the external resource is successfully converted into the standard resource supported by the resource publishing platform.

[0122] Among them, in an illustrated embodiment, since multiple subtasks included in the above conversion task may utilize different inference capabilities of the LLM model; therefore, in order to make full use of the differences in the inference capabilities of different LLM models, each subtask included in the above conversion task can be executed using different LLM models.

[0123] It should be noted that the trigger conditions for the events used to trigger the execution of the subtasks included in the above conversion task are not particularly limited in this specification. In practical applications, the trigger conditions for the event can be flexibly defined based on specific requirements.

[0124] In an illustrated embodiment, the trigger condition for the event used to trigger the execution of any subtask included in the above conversion task may specifically include that the execution result of the previous subtask of the subtask has been confirmed by the user.

[0125] For example, taking the above first event as an example, since the first event is generated after the first subtask is executed and is used to trigger the execution of the next second subtask of the first subtask, therefore, in this case, when determining whether the first event meets the trigger condition, it can specifically determine whether the user has confirmed the first execution result; if the user has confirmed the first execution result, it can be determined at this time that the first event meets the trigger condition.

[0126] Among them, the specific manner in which the user confirms the execution result of the subtask is not particularly limited in this specification.

[0127] For example, the execution result output by the LLM model can be output and displayed to the user through a preset display interface; specifically, a user option (such as a user button) for confirming the execution result can be included in the display interface, and the user can perform an operation on the user option in the display interface to manually confirm the execution result output by the LLM model.

[0128] In one illustrated embodiment, if the user does not confirm the execution result of a certain subtask included in the above conversion task, it indicates that the user is not satisfied with the output result of the subtask. In this case, a mechanism for iteratively optimizing the prompt words of the subtask can be introduced to guide the LLM model to output a satisfactory execution result for the user.

[0129] Among them, the mechanism for iteratively optimizing the prompt words of the subtask specifically refers to iteratively optimizing and adjusting the generated prompt words of the subtask based on a preset optimization adjustment strategy until the LLM model outputs a satisfactory execution result that allows the user to perform manual confirmation based on the optimized and adjusted prompt words.

[0130] For example, taking the first subtask included in the above conversion task as an example, if the user does not confirm the first execution result of the first subtask, at this time, the generated first prompt word is optimized and adjusted based on the preset optimization adjustment strategy, and the optimized and adjusted first prompt word is re-input into the LLM model so that the LLM model performs the inference calculation corresponding to the first subtask based on the first prompt word; then, the first execution result corresponding to the first subtask output by the LLM model can be obtained, and it is re-determined whether the user has confirmed the first execution result; if the user still does not confirm the first execution result, at this time, the first prompt word generated can be continuously optimized and adjusted based on the preset optimization adjustment strategy, and so on, and the above process is repeatedly executed until the user confirms the first execution result corresponding to the first subtask output by the LLM model.

[0131] Among them, it should be noted that the optimization adjustment strategy used for optimizing and adjusting the prompt words mentioned above is not specifically limited in this specification. In actual applications, it can be flexibly configured based on specific requirements.

[0132] In one illustrated embodiment, in order to avoid the situation that the LLM model still cannot output a satisfactory execution result for a long time after repeatedly iteratively optimizing and adjusting the prompt words, a mechanism for the user to manually modify the execution result output by the LLM model can also be introduced.

[0133] In this case, the above display interface may also include a user option (such as a dropdown list) for modifying the execution result, and the user can manually modify the execution result output by the LLM model by operating on this user option in the display interface. The above resource publishing platform can respond to the user operation on this user option, obtain the updated data input by the user corresponding to the execution result output by the LLM model, and then modify and replace the execution result based on this updated data.

[0134] It should be noted that the splitting method adopted when splitting the execution link of the above conversion task is not particularly limited in this specification. In actual applications, the above conversion task can be flexibly split based on specific requirements. Correspondingly, after the conversion task is split, the task objectives of the multiple subtasks included in the conversion task usually depend on the actual splitting method adopted and are not particularly limited in this specification.

[0135] The following is illustrated by a specific example of splitting the above conversion task.

[0136] Please refer to Figure 4 , Figure 4 which is a flowchart showing a method of converting external resources into standard resources in this specification.

[0137] As Figure 4 shown, in an illustrated embodiment, after splitting the above conversion task, the subtasks included in the above conversion task may specifically include Figure 4 the 4 subtasks shown, namely the resource allocation subtask (i.e., the first subtask), the resource matching subtask (i.e., the second subtask), the information extraction subtask (i.e., the third subtask), and the detail map production subtask (i.e., the fourth subtask).

[0138] The task objective of the above resource classification subtask is specifically used to determine the resource type of the external resource to be published;

[0139] The task objective of the above resource matching subtask is specifically used to match the resource information of the external resource with the resource information of the standard resource corresponding to the resource type of the external resource stored in the standard resource database, so as to establish a correspondence between the information items included in the resource information of the external resource and the standard information items included in the resource information of the standard resource;

[0140] The task objective of the above information extraction subtask is specifically used to extract the rule information related to the above standard resource from the resource information of the external resource, which is the third subtask;

[0141] The task objective of the above-mentioned detailed diagram generation subtask is specifically used to generate a standard resource detailed diagram corresponding to the external resource based on the resource information of the external resource, the above-established corresponding relationship, and the above-extracted rule information.

[0142] Please continue to refer to Figure 4 , in Figure 4 the process shown, an event-driven task execution framework shown in Figure 3 can still be adopted, and the LLM model is driven to execute the multiple subtasks in a way triggered by events.

[0143] In addition, during the process of driving the LLM model to execute the multiple subtasks through events, whether the user has confirmed the execution result of a certain subtask that has been executed can also be used as the triggering condition for the event of the next subtask that drives the subtask. At the same time, a mechanism for iteratively optimizing the prompt words for each subtask can be introduced to guide the LLM model to output an execution result satisfactory to the user, and then apply the execution result.

[0144] In this case, when using the above LLM model to execute the above conversion task, first, based on the resource information of the external resource to be published obtained, a first prompt word corresponding to the above resource classification subtask can be generated, and the first prompt word is input into the LLM model so that the LLM model performs inference calculations corresponding to the first subtask based on the first prompt word to determine the resource type of the external resource.

[0145] For example, in actual applications, the resource information of the external resource and the category specifications defined by the platform corresponding to the standard resource can be used as the content of the first prompt word to generate the first prompt word. And the LLM model can match the resource information of the external resource and the category specifications included in the prompt word to clarify the resource type of the external resource.

[0146] After the above resource classification subtask is completed, the resource type of the external resource output by the LLM model can be output and displayed to the user through a preset display interface.

[0147] Among them, the display interface can specifically include a user option (such as a user button) for confirming the execution result. And the user can manually confirm the execution result output by the LLM model by operating the user option in the display interface.

[0148] For example, taking the ticket products provided by the above-mentioned resources as external product providers as an example, the product information of the above-mentioned external products to be obtained may specifically include information such as the name of the ticket product, the name of the scenic spot, the country where the scenic spot is located, the city where the scenic spot is located, the address of the scenic spot, and the cost of the scenic spot. The product type of the external product determined by the LLM model based on this information can be output and displayed on this display interface. In addition, the classification reason for the LLM model to classify the external product can also be output and displayed on this display interface for users to refer to. When the user confirms that the product type of the external product is the correct product type supported by the product release platform, the above user options can be operated to manually confirm the execution result output by the LLM model.

[0149] Of course, if the product type of the external product output by the LLM model is not the correct product type, at this time, the above mechanism for iteratively optimizing the prompt words can be used to iteratively optimize the above first prompt word until the LLM model outputs a satisfactory execution result that allows the user to perform manual confirmation based on the optimized prompt word.

[0150] After the above resource classification subtask is completed, the resource type of the external resource output by the LLM model can be continuously obtained, and a first event for triggering the execution of the above resource matching subtask can be generated; then, it can be determined whether the user has confirmed the resource type of the external resource output by the LLM model; if so, at this time, the first event can be executed, and further based on the resource information of the external resource and the resource type of the external resource, a second prompt word corresponding to the above resource matching subtask can be generated;

[0151] For example, the resource information of the external resource and the resource type of the external resource can be used as the content of the second prompt word to generate the second prompt word.

[0152] Then, the second prompt word can be input into the LLM model, so that the LLM model performs inference calculations corresponding to the above resource matching subtask based on the second prompt word, and matches the resource information of the external resource with the resource information of the standard resource corresponding to the resource type stored in the standard resource database, so as to establish a corresponding relationship between the information items included in the resource information of the external resource and the standard information items included in the resource information of the standard resource.

[0153] Of course, if the above corresponding relationship output by the LLM model is not the correct corresponding relationship, at this time, the above mechanism for iteratively optimizing the prompt words can be used to iteratively optimize the above second prompt word until the LLM model outputs a satisfactory execution result that allows the user to perform manual confirmation based on the optimized prompt word.

[0154] Among them, in one of the illustrated embodiments, the resource information of the above external resource may specifically include multiple information items; in this case, multiple matching prompt words corresponding to the multiple information items can be generated based on the resource information of the external resource and the resource type of the external resource.

[0155] Then, the multiple generated matching prompt words can be respectively input into the LLM model, so that the LLM model performs inference calculations corresponding to the above resource matching subtask based on the input matching prompt words, and matches the information items corresponding to the matching prompt words included in the resource information of the external resource with the standard information items included in the resource information of the standard resource corresponding to the resource type stored in the standard resource database, and establishes a corresponding relationship between the information item and the standard information item based on the matching result.

[0156] For example, still taking the above resource as a ticket-type product provided by an external product provider as an example, the product information of the above external product may specifically include information items such as scenic spots, names of charging items, ticket types, show times, and regions. However, since these information items included in the product information of the external product usually do not conform to the information specifications of the above product release platform for standard products, it is necessary to match the information items such as scenic spots, names of charging items, ticket types, show times, and regions included in the product information of the external product with the standard information items of the ticket-type products stored in the standard product database respectively, in order to establish a corresponding relationship between these non-standard information items included in the product information of the external product and the standard information items of the ticket-type products stored in the standard product database.

[0157] Among them, in one of the illustrated embodiments, before generating the second prompt word corresponding to the above resource matching subtask, the similarity between the information items included in the resource information of the external resource and the standard information items included in the resource information of the standard resource corresponding to the resource type of the external resource stored in the standard resource database can also be calculated in advance, and at least one standard information item with the highest similarity to the information items included in the resource information of the external resource stored in the standard resource database can be obtained.

[0158] For example, in one example, the above similarity can be specifically represented by a vector distance. In this case, when calculating the similarity between the information items included in the resource information of the external resource and the standard information items included in the resource information of the standard resource corresponding to the resource type of the external resource stored in the standard resource database, specifically, the information items included in the resource information of the external resource can be converted into an information vector, and the standard information items included in the resource information of the standard resource corresponding to the resource type of the external resource stored in the standard resource database can also be converted into a standard information vector. Then, the vector distance between the information vector and the standard information vector can be calculated, and at least one standard information vector with the largest vector distance between the information vectors corresponding to the information items included in the resource information of the external resource stored in the standard resource database can be determined. Finally, at least one standard information item corresponding to at least one of the above standard information vectors with the largest vector distance stored in the standard resource database can be obtained as at least one of the above standard information items with the highest similarity.

[0159] In this case, when generating the second prompt corresponding to the above resource matching subtask, specifically, based on the resource information of the external resource and at least one of the above standard information items with the highest similarity pre-screened, the second prompt corresponding to the above resource matching subtask can be generated.

[0160] For example, the resource information of the external resource and at least one of the above standard information items with the highest similarity screened out can be used as the content of the second prompt to generate the second prompt.

[0161] Then, the second prompt can be input into the LLM model, so that the LLM model performs the inference calculation corresponding to the above resource matching subtask based on the second prompt, matches the information items included in the resource information of the external resource with the above at least one standard information item included in the prompt respectively, and establishes the corresponding relationship between the information items included in the resource information of the external resource and the above at least one standard information item based on the matching result.

[0162] In this way, at least one of the above standard information items with the highest similarity pre-screened can be directly added to the prompt, clearly defining the matching range in advance for the LLM model, which can help improve the accuracy of the matching result output by the LLM model.

[0163] After the above resource matching subtask is completed, the above corresponding relationship output by the LLM model can be continuously obtained, and a second event for triggering the execution of the above information extraction subtask can be generated. Then, it can be determined whether the user has confirmed the above corresponding relationship. If so, at this time, the second event can be executed, and a third prompt corresponding to the above information extraction subtask can be further generated based on the resource information of the external resource.

[0164] For example, the resource information of the external resource can be used as the content of the third prompt word to generate the third prompt word.

[0165] Then, the third prompt word can be input into the LLM model, so that the LLM model performs inference calculations corresponding to the above information extraction subtask based on the third prompt word, and extracts rule information related to the standard resource from the resource information of the external resource.

[0166] Of course, if the above rule information output by the LLM model is not the correct rule information, at this time, the above mechanism for iteratively optimizing the prompt word can also be used to iteratively optimize the above third prompt word until the LLM model outputs a satisfactory execution result that allows the user to perform manual confirmation based on the optimized and adjusted prompt word.

[0167] Among them, the rule information can specifically include any form of rules related to the standard resource.

[0168] For example, still taking the above-mentioned resource as the commodity provided by the external commodity provider as an example, the above rule information can include commodity rules; for example, taking the above-mentioned commodity as a ticket commodity as an example, the above commodity rules can specifically include any one or more of the following combinations: reservation rules, cancellation rules, refund rules, purchase limit rules, usage rules.

[0169] After the above information extraction subtask is completed, the above rule information output by the LLM model can be continuously obtained, and a third event for triggering the execution of the above detailed diagram making subtask can be generated; then, it can be determined whether the user has confirmed the above rule information extracted; if so, at this time, the third event can be executed, and further based on the resource information of the external resource, the above-established corresponding relationship (that is, the execution result of the above resource matching subtask), and the above-extracted rule information (that is, the execution result of the above extraction subtask), a fourth prompt word corresponding to the above detailed diagram making subtask can be generated.

[0170] For example, the resource information of the external resource, the above-established corresponding relationship, and the above-extracted rule information can be used as the content of the fourth prompt word to generate the fourth prompt word.

[0171] Then, the fourth prompt word can be input into the LLM model, so that the LLM model performs inference calculations corresponding to the fourth subtask based on the fourth prompt word, and uses the image generation ability of the LLM model to generate a standard resource detailed diagram corresponding to the above external resource based on the resource information of the external resource, the above-established corresponding relationship, and the above-extracted rule information.

[0172] Among them, the process of generating a standard resource details map corresponding to the above external resource by the LLM model using its own image generation ability based on the resource information of the external resource, the established corresponding relationship, and the extracted rule information will not be elaborated in this specification.

[0173] For example, in an example, still taking the ticket-type product provided by the above external product provider as an example, when generating a standard product details map based on the product information of the ticket-type product, the established corresponding relationship, and the extracted rule information, in one case, the standard information items included in the corresponding relationship can be used as keys, and the non-standard information items corresponding to the standard information items included in the product information of the ticket-type product can be used as the values corresponding to the keys, and they can be associated and displayed in the generated standard resource details map; or, in another case, the non-standard information items included in the product information of the ticket-type product can be directly replaced with the standard information items corresponding to the non-standard information items, so as to convert the product information included in the ticket-type product into a standard product that conforms to the information format supported by the above product release platform. For the above rule information, since it is itself a product rule related to the standard product, these product rules can be directly displayed in the standard resource details map.

[0174] Among them, it should be noted that the specific layout of various information included in the finally generated standard resource details map on the details map will not be specifically limited in this specification, and in actual applications, it can be flexibly designed according to specific requirements.

[0175] When the LLM model uses its own image generation ability to generate a standard resource details map corresponding to the above external resource based on the resource information of the external resource, the established corresponding relationship, and the extracted rule information, due to the obvious difference in the richness of the resource information of different external resources, if a unified standard resource details map structure is used to generate a unified standard resource details map, there may be a problem that a standard resource details map that can reasonably display the resource information of the external resource cannot be generated.

[0176] For example, if the structure of the standard resource details map is too complex, there may be a problem that a lot of information on the details map is missing; if the structure of the standard resource details map is too simple, it is difficult to reflect the real information of the external resource.

[0177] Based on this, in one of the illustrated embodiments, when using the image generation ability of the LLM model to generate a standard resource detail map corresponding to the external resource based on the resource information of the external resource, the established corresponding relationship, and the extracted rule information, instead of using a unified standard resource detail map structure, a standard resource detail map with a higher degree of freedom can be adaptively generated for the external resource based on the amount of information actually contained in the resource information of the external resource. That is, for different external resources, instead of generating standard resource detail maps with exactly the same structure, standard resource detail maps that can cover the resource information of the external resource are generated based on the amount of information actually contained in different external resources.

[0178] In this way, a standard resource detail map that better meets the user's needs can be generated for the user.

[0179] For example, taking the above resource as a commodity, instead of using a unified structure, a commodity detail map with a higher degree of freedom is adaptively generated for the external commodity based on the amount of information actually contained in the external commodity. Obviously, it can stimulate the user's desire to purchase and reduce the communication costs before and after sales.

[0180] Step 106: Obtain the standard resource output by the LLM model and publish the standard resource on the resource publishing platform.

[0181] After converting the above external resource into a standard resource based on the LLM model, the resource publishing platform can obtain the converted standard resource output by the LLM model, and then publish the standard resource on the resource publishing platform.

[0182] For example, the standard resource detail map corresponding to the external resource output by the LLM model can be published on the resource publishing platform.

[0183] In one of the illustrated embodiments, the LLM model accessed by the resource publishing platform can specifically be a general LLM model; a task execution framework based on the general LLM model can specifically be carried on the resource publishing platform.

[0184] Among them, the task execution framework can specifically adopt a modular hierarchical structure. Through clear hierarchical division and functional modularization, the task execution framework can have the characteristics of high efficiency, strong scalability, and flexibility. In the task execution framework adopting the modular hierarchical structure, the multiple subtasks included in the conversion task are usually functional modules defined by the user in the task execution framework.

[0185] Please refer to Figure 5 , Figure 5An architecture diagram of a task execution framework based on a general LLM model shown in this specification;

[0186] As Figure 5 shown, the task execution framework can specifically adopt a modular hierarchical structure, which can specifically include an interface layer, a data processing layer, a task processing layer, a function module layer, and a bottom support layer. In addition, the task execution framework can also provide an independent tool module for users.

[0187] Among them, the above-mentioned interface layer is specifically used to provide API interfaces related to task execution.

[0188] As Figure 5 shown, the interfaces provided by the interface layer can specifically include APIs such as those for viewing the task list, viewing task details, triggering task execution, manually modifying the output interface of the LLM model, and viewing the task execution results of the LLM model output, etc. When the user executes the above conversion task based on this task execution framework, these API interfaces provided by the interface layer can be called to implement service functions related to task execution.

[0189] For example, as Figure 5 shown, when the user is not satisfied with the execution result output by the LLM model, the user can operate on the display interface for displaying the execution result to call the above-mentioned API for manually modifying the output interface of the LLM model, and then manually modify the execution result output by the LLM model in this display interface.

[0190] The above-mentioned data processing layer is specifically used to provide data processing capabilities related to task execution.

[0191] As Figure 5 shown, this data processing layer can specifically provide the ability to preprocess data related to task execution, the ability to convert data related to task execution, and the ability to store data related to task execution, etc.

[0192] For example, as mentioned in the previous embodiments, before generating the second prompt corresponding to the resource matching subtask, it is also possible to pre-compute the similarity between the information items included in the resource information of the external resource and the standard information items included in the resource information of the standard resource corresponding to the resource type of the external resource stored in the standard resource database, and obtain at least one standard information item with the highest similarity to the information items included in the resource information of the external resource stored in the standard resource database. This processing operation utilizes the data preprocessing ability provided by the data processing layer. The processing operation of converting the information items included in the resource information of the external resource into information vectors, as mentioned in the previous embodiments, utilizes the data conversion ability provided by the data processing layer.

[0193] The above task processing layer can specifically be used to provide task execution functions related to task execution.

[0194] Such as Figure 5 As shown, the task execution functions related to task execution provided by the above task processing layer can specifically include a task scheduling function for task scheduling, a task execution engine for executing tasks, and a task status management function for managing the status of tasks, and so on.

[0195] The above function module layer is the core layer of the above task execution framework and is used to provide function modules corresponding to multiple subtasks included in the above conversion task.

[0196] In practical applications, when following the process of manually executing the conversion task and splitting the execution link of the conversion task into multiple subtasks, based on the extension ability of the function module layer, function modules corresponding to each subtask included in the above conversion task can be defined on the function module. Thus, through the function module layer, the above task execution framework can support different types of function modules and quickly access new function modules by utilizing the extension ability of the function module layer.

[0197] When the execution link of the conversion task is split into multiple subtasks and the execution results of the multiple subtasks do not meet the user's expectations, the user can also optimize and adjust the execution link of the conversion task, re-split the conversion task, and then define function modules corresponding to each re-split subtask on the function module, and so on. By iteratively executing the above optimization process, it is possible until the execution results of the multiple split subtasks can meet the user's expectations.

[0198] Such as Figure 5As shown in the figure, taking the splitting of the above conversion task into the above resource allocation subtask, resource matching subtask, information extraction subtask, and detailed map production subtask as an example, corresponding function modules can be defined for these four types of subtasks on this function module layer.

[0199] The above-mentioned underlying support layer is used to provide some underlying supports related to task execution.

[0200] Such as Figure 5 As shown in the figure, the underlying supports provided by the above-mentioned underlying support layer can specifically include SQL, configuration management, vector database, and so on.

[0201] The above-mentioned tool module can specifically be used to provide some daily tools related to this task execution framework.

[0202] Such as Figure 5 As shown in the figure, the daily tools provided by the above-mentioned tool module can specifically include log management tools, performance monitoring tools, cost assessment tools, result sampling inspection tools, cache tools, prompt word maintenance tools, and so on.

[0203] Based on the above task execution framework, a complete supply chain can be provided to convert the external resources provided by external resource providers into standard resources and publish them on this resource publishing platform. When the resource publishing platform needs to publish the external resources provided by external resource providers, there is no need for manual participation throughout the process. After remotely pulling the resource information of the external resources from the external resource provider through API calls, it can automatically convert the external resources into standard resources and publish them on this resource publishing platform, thereby significantly reducing the labor cost in publishing external resources and improving the publishing efficiency of external resources.

[0204] For example, still taking the above resources as ticket products provided by an external commodity provider and the above resource publishing platform as a travel trading service platform as an example, based on the above task execution framework, a complete commodity supply chain can be provided to convert the ticket products provided by the external commodity provider into standard ticket products with the information format supported by this travel trading service platform and publish them on this resource publishing platform. When the travel trading service platform needs to publish the ticket products provided by the external commodity provider, there is no need for manual participation throughout the process. After remotely pulling the commodity information of the ticket products (such as the name of the ticket product, the name of the scenic spot, the country where the scenic spot belongs, the city where the scenic spot belongs, the address of the scenic spot, the cost of the scenic spot, etc.) from the external commodity provider through API calls, it can automatically convert the ticket products into standard ticket products and sell them on this travel trading service platform.

[0205] In the above technical solution, by introducing an LLM model in the resource publishing platform to perform the conversion task of converting external resources into standard resources supported by the resource publishing platform, when publishing external resources, prompt words can be generated to guide the LLM model to convert external resources into standard resources, so that external resources can be automatically converted into standard resources for publishing without manual participation, thereby reducing labor costs and improving the publishing efficiency of external resources.

[0206] Corresponding to the embodiments of the foregoing method, this specification also provides embodiments of a device, an electronic device, and a storage medium.

[0207] Figure 6 It is a schematic structural diagram of an electronic device provided by an exemplary embodiment. Please refer to Figure 6 , at the hardware level, the device includes a processor 602, an internal bus 604, a network interface 606, a memory 608, and a non-volatile memory 610. Of course, other required hardware may also be included. One or more embodiments of this specification can be implemented in a software manner. For example, the processor 602 reads the corresponding computer program from the non-volatile memory 610 into the memory 608 and then runs it. Of course, in addition to the software implementation manner, one or more embodiments of this specification do not exclude other implementation manners, such as a logic device or a combination of software and hardware, etc. That is, the execution subject of the following processing flow is not limited to each logic unit, and can also be hardware or a logic device.

[0208] As Figure 7 shown, Figure 7 is a block diagram of a resource publishing device based on an LLM model shown in accordance with an exemplary embodiment of this specification. The device can be applied to an electronic device as shown in Figure 6 to implement the technical solution of this specification. Among them, the resource publishing platform performs the conversion task of converting external resources into standard resources supported by the resource publishing platform based on the accessed LLM model; the device 1200 includes:

[0209] An acquisition module 701 that acquires the resource information of the external resource to be published;

[0210] A conversion module 702 that generates a prompt word corresponding to the conversion task based on the resource information of the external resource, and inputs the prompt word into the LLM model, so that the LLM model performs inference calculations corresponding to the conversion task based on the prompt word to convert the external resource into the standard resource;

[0211] A publishing module 703 that acquires the standard resource output by the LLM model and publishes the standard resource on the resource publishing platform.

[0212] Correspondingly, this specification also provides an electronic device, which includes a processor and a memory for storing instructions executable by the processor. Wherein, the processor is configured to implement the steps in all the method flows described above.

[0213] Correspondingly, this specification also provides a computer-readable storage medium, on which executable computer program instructions are stored. Wherein, when the instructions are executed by a processor, the steps in all the method flows described above are implemented.

[0214] Correspondingly, this specification also provides a computer program product, on which executable computer program instructions are stored. Wherein, when the computer program instructions are executed by a processor, the steps in all the method flows described above are implemented.

[0215] The systems, devices, modules or units illustrated in the above embodiments can be specifically implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a server system. Of course, with the development of future computer technology, it is not excluded that a computer for implementing the functions of the above embodiments can be, for example, a personal computer, a laptop computer, an in-vehicle human-machine interaction device, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0216] Although one or more embodiments of this specification provide method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-creative means. The order of steps listed in the embodiments is only one way among many execution orders of steps and does not represent the only execution order. When an actual device or terminal product is executing, it can be executed in the order of the method shown in the embodiments or the drawings or executed in parallel (for example, in a parallel processor or multi-threaded processing environment, or even in a distributed data processing environment). The term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, product or device comprising a series of elements includes not only those elements but also other elements not explicitly listed, or also includes elements inherent to such process, method, product or device. Without further limitation, it does not exclude the presence of additional identical or equivalent elements in the process, method, product or device comprising the said elements. For example, if the terms first, second, etc. are used to denote names, they do not denote any specific order.

[0217] For the convenience of description, when describing the above device, various modules are described separately according to their functions. Of course, when implementing one or more of this specification, the functions of each module can be implemented in the same or multiple software and / or hardware, or the modules implementing the same function can be realized by a combination of multiple sub-modules or sub-units, etc. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the couplings or direct couplings or communication connections shown or discussed with each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical or other forms.

[0218] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0219] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0220] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0221] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and a memory.

[0222] Memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM) and / or non-volatile memory such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0223] Computer-readable media includes both permanent and non-permanent, removable and non-removable media implemented by any method or technology for information storage. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile discs (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage, graphene storage or other magnetic storage devices, or any other non-transitory media that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media such as modulated data signals and carrier waves.

[0224] Those skilled in the art will appreciate that one or more embodiments of this specification may be provided as a method, system, or computer program product. Accordingly, one or more embodiments of this specification may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, one or more embodiments of this specification may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0225] One or more embodiments of this specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. One or more embodiments of this specification may also be practiced in a distributed computing environment where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including storage devices.

[0226] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other, and the key point of each embodiment is to illustrate the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and for the relevant parts, reference can be made to the description of the method embodiment. In the description of this specification, the description of reference terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of this specification. In this specification, the schematic expression of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0227] The above description is only for the embodiments of one or more embodiments of this specification and is not intended to limit one or more embodiments of this specification. For those skilled in the art, one or more embodiments of this specification can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this specification shall be included within the scope of the claims.

Claims

1. A resource publishing method based on an LLM model, applied to a resource publishing platform, wherein the resource publishing platform performs a conversion task of converting external resources into standard resources supported by the resource publishing platform based on the accessed LLM model; the method comprises: Acquire multiple resource indicators of each external resource in the external resource set provided by the external resource provider; wherein the multiple resource indicators are indicators used to represent the resource value of the external resource; the multiple resource indicators include a first indicator used to represent the flow value of the external resource, a second indicator used to represent the transaction value of the external resource, and a third indicator used to represent the change frequency of the resource information of the external resource; The first indicator, the second indicator and the third indicator are weightedly calculated to obtain a resource value score of each external resource in the external resource set, and the calculated resource value score is further mapped into a scheduling priority; wherein the resource value score represents the resource value of the external resource; and the scheduling priority is positively correlated with the resource value score; Periodically performing resource scheduling on the external resources in the external resource set based on the scheduling priority, so as to determine the external resources to be released from the external resource set, and obtaining resource information of the determined external resources to be released from the external resource provider; Based on the resource information of the external resource, a prompt word corresponding to the conversion task is generated, and the prompt word is input into the LLM model, so that the LLM model performs an inference calculation corresponding to the conversion task based on the prompt word to convert the external resource into the standard resource; The standard resource output by the LLM model is obtained, and the standard resource is published on the resource publishing platform.

2. The method according to claim 1, wherein the first indicator comprises the ratio of the number of visits to the external resource to the total number of visits to each external resource in the external resource set; the second indicator comprises the product of the number of orders for the external resource and the order amount of the external resource; the third indicator comprises a weighted sum obtained by weighted calculation of the number of times each information field contained in the resource information of the external resource changes within a preset time period in the future; in, The weighting coefficients of the first indicator and the second indicator are weighting coefficients calculated based on the resource conversion rate of the external resources; the weighting coefficient of the third indicator is a weighting coefficient calculated based on the order information of orders for failed transactions for the external resources due to changes in resource information.

3. The method of claim 1, further mapping the calculated resource value score into a scheduling priority, comprising: The calculated resource value score is further mapped to a scheduling priority based on the following expression: P= In this expression, P represents the scheduling priority of the external resource; S represents the resource value score of the external resource; min represents the minimum value of the resource value scores of all the external resources in the external resource set; max represents the maximum value of the resource value scores of all the external resources in the external resource set; g represents the number of priority levels preset for the scheduling priority.

4. The method according to claim 3, periodically performing resource scheduling on the external resources in the external resource set based on the scheduling priority to determine the external resources to be released from the external resource set, comprising: Creating a scheduling node corresponding to an external resource in the external resource set; wherein the information recorded in the scheduling node includes the scheduling priority of the external resource; Using the scheduling priority as a sorting key, constructing a binary tree of scheduling nodes corresponding to the external resources in the external resource set, and periodically traversing the binary tree to search for at least one target scheduling node with the highest scheduling priority; The external resource corresponding to the at least one found target scheduling node is determined as the external resource to be released.

5. The method according to claim 4, wherein the information recorded in the scheduling node further includes a scheduling weight value corresponding to the priority level of the scheduling priority of the external resource; wherein, The scheduling priority of the external resource is positively correlated with the scheduling weight value; Using the scheduling priority as a sorting key, constructing the scheduling nodes corresponding to the external resources in the external resource set into a binary tree, and periodically traversing the binary tree to find at least one target scheduling node with the highest scheduling priority, including: Using the scheduling weight value as a sorting key, the scheduling nodes corresponding to the external resources in the external resource set are constructed into a binary tree, and the binary tree is periodically traversed to find at least one target scheduling node with the highest scheduling weight value.

6. The method of claim 1, wherein the conversion task comprises a plurality of subtasks executed based on event triggering; Based on the resource information of the external resource, a prompt word corresponding to the conversion task is generated, and the prompt word is input into the LLM model, so that the LLM model performs reasoning calculation based on the prompt word to convert the external resource into the standard resource, including: Based on the resource information of the external resource, generate a first prompt word corresponding to a first subtask included in the conversion task, and input the first prompt word into the LLM model so that the LLM model performs an inference calculation corresponding to the first subtask based on the first prompt word; Obtaining a first execution result corresponding to the first subtask output by the LLM model, and generating a first event for triggering the execution of a second subtask included in the conversion task; determining whether the first event satisfies a trigger condition; if the first event satisfies the trigger condition, executing the first event, further generating a second prompt word corresponding to the second subtask based on the first execution result, and inputting the second prompt word into the LLM model, so that the LLM model performs an inference calculation corresponding to the second subtask based on the second prompt word; Continue to obtain the second execution result corresponding to the second subtask, and repeat the above process until all subtasks included in the conversion task are executed, and the external resource is converted into the standard resource.

7. The method according to claim 6, wherein the triggering condition for triggering the event of executing any subtask included in the conversion task comprises that the execution result of the previous subtask of the subtask is confirmed by the user; Determining whether an event for triggering the execution of any subtask included in the conversion task satisfies a triggering condition includes: determining whether the user has confirmed the execution result of the previous subtask of any subtask included in the conversion task; If the user confirms the execution result of the previous subtask, it is determined that the event meets the trigger condition.

8. The method of claim 7, further comprising: If the user has not confirmed the execution result of the previous subtask, the generated prompt word corresponding to the previous subtask is optimized and adjusted based on a preset optimization and adjustment strategy, and the optimized and adjusted prompt word is re-input into the LLM model, so that the LLM model performs the reasoning calculation corresponding to the previous subtask based on the prompt word; and, the execution result corresponding to the previous subtask output by the LLM model is obtained, and it is determined whether the user has confirmed the execution result; the above process is repeated until the user confirms the execution result.

9. The method of claim 7, wherein the plurality of subtasks comprises: A first subtask for determining a resource type of the external resource; A second subtask for matching the resource information of the external resource with the resource information of the standard resource corresponding to the resource type stored in the standard resource database to establish a correspondence between the information items included in the resource information of the external resource and the standard information items included in the resource information of the standard resource; A third subtask for extracting rule information related to the standard resource from the resource information of the external resource; A fourth subtask is used to generate a standard resource detail map corresponding to the external resource based on the resource information of the external resource, the established corresponding relationship, and the extracted rule information.

10. According to the method as claimed in claim 9, the resource publishing platform includes a product publishing platform; the external resources include products provided by external product providers; the standard resources include products in an information format supported by the product publishing platform; and the rule information includes product rules.

11. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 10 are implemented.

12. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 10 are implemented.

Images