Information processing method and apparatus, storage medium, and electronic device

Abstract

The present disclosure provides an information processing method and device, a storage medium and an electronic device, and relates to the technical field of artificial intelligence, in particular to the technical field of deep learning, large model, intelligent store building, etc. The specific implementation scheme is: obtaining first material information of an intelligent store; based on the first material information, generating second material information of the intelligent store, the first material information and the second material information being used to build the intelligent store. In the embodiment of the present disclosure, the generative scheme can generate the second material information required for store building, thereby reducing the amount of information entered by the user, shortening the operation path of the user, and improving the store building efficiency of the user.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of artificial intelligence, in particular to the technical field of deep learning, large model, intelligent store building, etc. BACKGROUND

[0002] With the continuous evolution of Internet technology, online shopping has gradually become a mainstream consumption lifestyle. Traditional offline stores have also established online stores on the Internet with the development of Internet technology. Users need to deal with increasingly complex challenges when establishing online stores. One of the core problems is the need to submit tedious store materials and perform complex operation steps. It is also more difficult for small and medium-sized users or users lacking relevant knowledge and experience to be familiar with the store building method. SUMMARY

[0003] The present disclosure provides an information processing method, device, storage medium and electronic equipment.

[0004] According to an aspect of the present disclosure, an information processing method is provided, comprising:

[0005] obtaining first material information of an intelligent store;

[0006] generating second material information of the intelligent store based on the first material information;

[0007] The first material information and the second material information are used to build the intelligent store.

[0008] According to another aspect of the present disclosure, an information processing device is provided, comprising:

[0009] a first obtaining module configured to obtain first material information of an intelligent store;

[0010] a generating module configured to generate second material information of the intelligent store based on the first material information;

[0011] The first material information and the second material information are used to build the intelligent store.

[0012] According to another aspect of the present disclosure, an electronic equipment is provided, comprising:

[0013] at least one processor; and

[0014] a memory communicatively connected to the at least one processor; wherein

[0015] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method of any embodiment of the present disclosure.

[0016] According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the method according to any of the embodiments of the present disclosure.

[0017] According to another aspect of the present disclosure, there is provided a computer program product comprising a computer program which, when executed by a processor, implements the method according to any of the embodiments of the present disclosure.

[0018] The second material information required for store opening can be generated by the generative scheme in the embodiments of the present disclosure, which can reduce the amount of information entered by the user, shorten the operation path of the user, and improve the store opening efficiency of the user.

[0019] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present disclosure, nor to limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0020] The accompanying drawings are used to better understand the present scheme and do not limit the present disclosure. Among them:

[0021] Figure 1 is a flowchart of an information processing method according to an embodiment of the present disclosure;

[0022] Figure 2 is a flowchart of a constituting atom generation capability according to an embodiment of the present disclosure;

[0023] Figure 3 is a flowchart of a store name generation capability according to an embodiment of the present disclosure;

[0024] Figure 4 is a flowchart of a product label generation capability according to an embodiment of the present disclosure;

[0025] Figure 5 is a flowchart of a multimedia resource generation capability according to an embodiment of the present disclosure;

[0026] Figure 6 is a schematic diagram of a style example according to an embodiment of the present disclosure;

[0027] Figure 7 is a schematic diagram of a color distribution example according to an embodiment of the present disclosure;

[0028] Figure 8a is a flowchart of an existing generative intelligent store according to an embodiment of the present disclosure;

[0029] Figure 8bis a flowchart of generating an intelligent store according to an embodiment of the present disclosure;

[0030] Figure 9a is a page schematic diagram of generating an intelligent store according to an embodiment of the present disclosure;

[0031] Figure 9b is another page schematic diagram of generating an intelligent store according to an embodiment of the present disclosure;

[0032] Figure 9c is another page schematic diagram of generating an intelligent store according to an embodiment of the present disclosure;

[0033] Figure 9d is a store homepage view and a marketing page view generated by an intelligent website building system according to an embodiment of the present disclosure;

[0034] Figure 9e is a store homepage view and a marketing page view adopted by a target object according to an embodiment of the present disclosure;

[0035] Figure 10 is a structural schematic diagram of an atomic generation capability module assembled in an embodiment of the present disclosure;

[0036] Figure 11 is a structural schematic diagram of an information processing device according to an embodiment of the present disclosure;

[0037] Figure 12 is a block diagram of an electronic device for implementing an information processing method according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0038] Exemplary embodiments of the present disclosure are described herein with reference to the accompanying drawings, in which various details of the present disclosure are set forth to facilitate an understanding, and should be considered in a descriptive sense. It will be readily apparent to one of ordinary skill in the art that various changes and modifications of the embodiments described herein can be made without departing from the scope of the present disclosure. Likewise, the present disclosure intends to embrace all alternatives, modifications, and variations of described embodiments that fall within the scope of the present disclosure. Likewise, to avoid obscuring the present disclosure, the description omits well-known functions and constructions.

[0039] In addition, the terms "first", "second", etc. are used only for the purpose of description, and should not be understood as indicating or implying relative importance or implying the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present disclosure, the meaning of "a plurality of" is two or more, unless otherwise specifically limited.

[0040] With the rapid development of network technology, online shopping has gradually entered the lives of more and more people. Large and small online shopping systems, various types of intelligent stores can be seen everywhere in the process of browsing the Internet. The so-called intelligent store can be understood as: the intelligent store combines big data, intelligent software and hardware, and realizes the internetization, dataization and electrification of consumption management and marketing services of the store through management methods.

[0041] The existing intelligent store usually adopts automatic equipment and intelligent system to improve shopping efficiency and consumer experience. Consumers can easily select goods, check out and pay through the mobile application program or code scanning device of the intelligent store without queuing. At the same time, the intelligent store can also recommend personalized goods and preferential activities according to the purchase history and preferences of consumers, and provide more accurate shopping suggestions.

[0042] However, the existing intelligent store building system has high operation cost for the target object (such as a merchant) who needs to be settled. The existing intelligent store building system usually requires the target object to fill in structured and information-rich form items to meet the requirements of store building materials and data structure of the system, so as to complete the store building. The target object needs to manually enter all the information to obtain the initial effect of the intelligent store, and then optimize the intelligent store by self-adjustment and decoration. The target object needs to actively participate in the whole process, and the operation is complicated, especially for novice users. This is because all the store building materials need to be manually provided, and the requirements of the store building materials cannot be well understood. Therefore, if the intelligent store building system used for store building can automatically generate relevant store building materials meeting the requirements, the complexity of store building can be reduced, and the store building process can be simplified.

[0043] In order to improve the automatic information generation capability of the intelligent store building system, simplify the process and reduce the operation cost of the target object, the information processing method provided by the embodiments of the present disclosure is proposed. It should be noted that the acquisition, storage and application of user information involved in the technical solutions provided by the embodiments of the present disclosure comply with the relevant laws and regulations and do not violate public order and good customs.

[0044] The information processing method provided by the embodiments of the present disclosure focuses on the convenience and efficiency of the operation of the target object and the generation capability of further completing the store building materials, so as to generate the material information required by the intelligent store with both effect and efficiency in a systematic way, thereby reducing the complexity of the operation of the target object.

[0045] Therefore, the implementation process of the information processing method is as shown in Figure 1 , which includes:

[0046] S101, acquiring first material information of an intelligent store.

[0047] The first material information can be various information input by the target object, including store building information defined according to business scenarios and business service requirements. In the embodiments of the present disclosure, the first material information can include at least one of the following: business license identification information, enterprise name, industry information, product name, and the like. When the target object inputs information, the information can be input in the form of text or in the form of a picture. For information input in the form of a picture, a picture-text recognition technology can be used to extract the required store building material information. For example, an OCR (Optical Character Recognition) recognition technology is used to recognize the first material information.

[0048] In S102, second material information of the intelligent store is generated based on the first material information; wherein the first material information and the second material information are used to build the intelligent store.

[0049] It can be understood that the embodiments of the present disclosure can provide an atomic generation capability for generating the second material information required to build the intelligent store based on the known first material information. In implementation, the first material information can be understood as original data that can be directly obtained by general users (i.e., the target object) and is generally actively provided by the target object. In this way, the requirement for the user to input the first material information is low, and the user does not need to understand or understand less about the store building requirements of the first material information, thereby simplifying the store building process.

[0050] Correspondingly, the second material information can be understood as store building material automatically generated by combining the first material information with the corresponding store building requirements, i.e., store building material automatically generated by the intelligent store building system. The second material information needs to meet relatively complex store building requirements, and therefore, the automatic generation of the second material information can further simplify user operations and improve user experience. Taking a store name as an example, the store name needs to meet certain risk control strategy requirements, for example, the store name needs to include certain information, and the length of the name needs to be limited within a certain range. In the present disclosure, the store name can be automatically generated and meet the risk control strategy requirements. The automatic generation of the store name can avoid the situation that the store building name obtained due to random user operation does not meet the requirements, resulting in the interruption of store building. The automatically generated store building name is generated depending on the store building requirements, and the store building name can be generated at one time as much as possible without the need for heavy user participation. In this way, the operation of the user is simple, and the operation complexity is low. Of course, the embodiments of the present disclosure also allow the user to modify the automatically generated second material information to meet the requirements of the user. This part will be described in detail below.

[0051] To improve the automation level of the intelligent store building system, various atomic generation capabilities can be provided in the embodiments of the present disclosure to generate corresponding second material information. The atomic generation capability refers to the capability of integrating, analyzing and generating useful information from input information and being able to independently run and provide services.

[0052] Continuing to take the store name as an example, an atomic generation capability of generating a store name can be constructed. In this way, in the process of building a store, the atomic generation capability is encapsulated in the intelligent store building system, improving the automation level of the intelligent store building system and helping users quickly and accurately build an intelligent store.

[0053] In summary, the embodiments of the present disclosure can generate useful second material information from the first material information, so that the target object can additionally generate other store building materials in the process of building an intelligent store. In the process of building a store, the target object does not need to fill in complex forms, and does not need to manually input all information, which reduces the understanding ability of the target object for store building requirements, is friendly to novice users, thereby reducing the store building operation cost of the target object, simplifying the process of generating an intelligent store, and improving the efficiency of generating an intelligent store.

[0054] To further simplify user operations, in the embodiments of the present disclosure, a dialog-based manner can be used to guide the target object to input the required first material information. Specifically, the target object can be interacted with to obtain the first material information of the intelligent store expected by the target object. In the interaction process, the user can be guided to input the store building materials in the form of text, video or animation. Moreover, a simple dialog form can be used to let the user complete the store building using a simple selection operation.

[0055] It can be understood that, in the case where the target object needs to generate an intelligent store, the target object and the intelligent store building system perform dialog-based communication. In order to input the store building materials (such as basic information, qualification information and industry information) of the target object into the intelligent store building system under the guidance of the intelligent store building system. In the process of interaction between the intelligent store building system and the target object, page interaction can be used, or voice interaction can be used, or gesture interaction can be used. The embodiments of the present disclosure do not limit the interaction manner, and the interaction manner capable of simplifying user operations is applicable to the embodiments of the present disclosure.

[0056] In the embodiments of the present disclosure, the required first material information can be quickly obtained through interaction with the target object. Through guided interaction, the target object can input the first material information, and the doubt of the target object about the store building material information to be input is reduced as much as possible. The target object's misunderstanding of the format and content of the required store building materials is avoided as much as possible, the trial and error cost is reduced as much as possible, which helps to improve the store building efficiency, reduces unnecessary repetitive work, saves time and computer processing resources.

[0057] In addition, the store building materials need to be audited in the intelligent store building process, and the intelligent store can be built based on the store building materials after the audit is passed. The audit of the store building material information includes multiple audit stages. In the traditional scheme, each audit stage is performed in series. For example, after passing the first audit stage, the target object is allowed to fill in the form of the second audit stage, and then the second audit stage is completed. After that, the target object is allowed to fill in the form of the third audit stage under the condition that the second audit stage is passed. In this way, the serial audit stages affect the efficiency of store building. It is also complex for users to operate.

[0058] In order to improve the efficiency of store building and simplify user operation, in the embodiments of the present disclosure, the required material information can be summarized and grouped to facilitate the use of parallel audit mechanism to complete the audit operation of multiple groups of information in parallel. All audit links do not have to be completed in series. Specifically, in the process of interacting with the target object, each session has a session mark. After collecting the first material information input by the target object based on the session, the information group belonging to the target quality inspection link can be identified based on the session mark, and the information group is input into the target quality inspection link to complete the quality inspection operation.

[0059] It can be understood that after multiple rounds of dialogue with the target object, the store building material information required for generating an intelligent store is obtained through the construction of an AIGC (Artificial Intelligence Generated Conversation) session string. In the process of dialogue with the target object, each round of dialogue has a corresponding session ID (Identification, identification code).

[0060] The information obtained from each round of dialogue is analyzed and identified, and the corresponding grouping information of the corresponding group is extracted. The information of multiple groups can be sent to the corresponding quality inspection environment in parallel for quality inspection, thereby completing the audit operation of the store building material information of the corresponding group.

[0061] In other embodiments, the information group required by the corresponding quality inspection link may include store building material information that needs to be automatically generated by the intelligent store building system, such as the second material information. The audit operation of the corresponding link can be completed in different cases. For example, in the process of interacting with the target object, the first material information is extracted based on the dialogue mark analysis, and in the case that the information in the group is not complete and the second material information needs to be generated, the corresponding second material information can be generated based on the first material information. Then, the full amount of information in the group is quality inspected to complete the corresponding audit operation.

[0062] In another case, if grouping is based on the session ID, the resulting grouping information is comprehensive, and the information in the group can be subjected to quality inspection operations after the corresponding review operations are completed.

[0063] In addition, in the case where the first material information needs to be quality inspected before the second material information can be generated, the first material information can be obtained based on the session mark and sent to the review link for quality inspection. If the second material information generated needs to continue to complete the quality inspection operation of another quality inspection link, the quality inspection of the information group to which the second material information belongs is continued.

[0064] In implementation, there is no limitation on when the second material information is generated and how the information group is constructed. The process can be built according to the needs of the actual application scene to process the store building material information.

[0065] In any quality inspection link, if the quality inspection is not passed, the target object can continue the dialogue to update the store building material information. If an exception occurs at any step, the session ID is used to re-perform the round of dialogue to ensure the integrity of the intelligent store and the store building material generation. In the case of re-session, the target object can be prompted about the reason why the store building material quality inspection is not passed, and the target object can be guided to improve the store building material.

[0066] In the embodiments of the present disclosure, based on the session mark, the quality inspection operation on the corresponding information can be implicitly completed in the interaction process, without the need to complete the information quality inspection process in series. For the target object, the store building material information required can be submitted without awareness, and the related quality inspection operation of the store building material information can be completed in the background, simplifying the user operation and improving the store building efficiency.

[0067] In the embodiments of the present disclosure, based on the foregoing content, a plurality of atomic generation capabilities can be provided to generate the store building material. The information required for each atomic generation capability can be the same or different. In order to generate high-quality store building material, the advantages of large models can be used to construct atomic generation capabilities in implementation.

[0068] The large model in the embodiments of the present disclosure generally refers to a model with a large number of parameters in the field of machine learning and artificial intelligence, which usually requires a large amount of computing resources for training and running. Such models can include language models, image models, reinforcement learning models, etc., which are large in size and can handle more complex tasks and data.

[0069] The large model can generate useful store building material information based on the prompt information. Accordingly, the second material information of the intelligent store based on the first material information mentioned in the foregoing of the embodiments of the present disclosure can be specifically executed as follows:

[0070] Step A1, generating prompt information based on the first material information.

[0071] The prompt information is the prompt of the large model. By constructing the prompt, the large model can understand the task requirements and generate reasonable second material information.

[0072] As shown in Figure 2 The embodiment of the present disclosure can exemplarily provide three atomic generation capabilities, including store name generation capability, product label generation capability, and multimedia resource generation capability. Each atomic generation capability corresponds to a category of information. Correspondingly, the first material information can include at least one of first sub-information, second sub-information, and third sub-information. The first sub-information is used for store name generation capability, the second sub-information is used for product label generation capability, and the third sub-information is used for multimedia resource generation capability.

[0073] Correspondingly, it can be understood that in the case that the first material information is the first sub-information required for generating a store name, the second material information is the store name.

[0074] In the case that the first material information is the second sub-information required for generating a product label, the second material information is the product label.

[0075] In the case that the first material information is the third sub-information required for generating a multimedia resource, the second material information is the multimedia resource.

[0076] In the embodiment of the present disclosure, the first material information can be classified into sub-information for generating different store building material information. Based on different categories of sub-information, the prompt information can be constructed based on a small amount of information to intelligently generate the required second material information. Thus, the automation degree of intelligent store building is improved, the efficiency and accuracy of intelligent store generation are improved, and the user operation is simplified.

[0077] In order to further simplify the user operation, the embodiment of the present disclosure expects to use as little information as possible to construct a few-shot prompt. Through analysis and mining of big data, the first sub-information used to construct the few-shot prompt in the embodiment of the present disclosure includes at least one of the following: enterprise name of the intelligent store, industry information of the intelligent store. The second sub-information used to construct the few-shot prompt includes at least one of the following: industry information of the intelligent store, product name, product specification, picture, and text. The third sub-information used to construct the few-shot prompt includes at least one of the following: store name, product category, product label, and original image.

[0078] In the embodiments of the present disclosure, the first material information is divided into different subcategories, and different categories contain specific store-opening material contents in the category. The information content is clear, and the input information is organized. A small amount of original store-opening materials can improve the efficiency of information input by the target object during the store-opening process, reduce the workload and error rate of input information. At the same time, the amount of information required for each atomic generation capability is very limited and basically has no ambiguity, so that the target object is easy to operate. At the same time, as little information as possible makes the information required for each atomic generation capability very small, which improves the user experience of the target object, and can generate the corresponding second material information through effective and small amount of information, thereby improving the efficiency of intelligent store generation.

[0079] As described above, the prompt information adopts the few-shot prompt composition scheme to reduce the amount of information required and simplify the store-opening operation. Since the information required during the store-opening process has certain requirements, in the embodiments of the present disclosure, important information in the first material information is extracted, and different modules of the few-shot prompt are further constructed in combination with the risk control strategy requirements to generate the prompt information, and then the second material information with high availability is generated based on the prompt information. Of course, the composition of the prompt information in the present disclosure is not limited to this.

[0080] In addition, in the embodiments of the present disclosure, the prompt information required for generating the intelligent store can also be constructed with reference to the reference store, specifically including the following steps:

[0081] Step B1, searching for a reference store of the intelligent store.

[0082] Step B2, constructing a guide example based on the corresponding relationship between the first material information of the reference store and the update result of the second material information of the reference store.

[0083] Step B3, constructing the prompt information based on the first material information of the intelligent store and the guide example.

[0084] In the embodiments of the present disclosure, the intelligent store generated by each target object and the first material information and the second material information in the process of generating the intelligent store are recorded in the intelligent store-opening system. When the target object needs to generate an intelligent store, and there is a similar reference store in the intelligent store-opening system, the intelligent store-opening system will automatically match the reference store, and use the first material information of the reference store recorded in the intelligent store-opening system and the corresponding second material information to construct the guide example of the current target object. That is, the current target object can use the store-opening material information of the reference store to construct the prompt information to guide the large model to quickly generate the second material information required by the current target object.

[0085] The embodiment of the present disclosure uses a reference store as an example guide, which is equivalent to using a successful case to build prompt information of the target object, reduces the operation complexity of the target object, and can accurately guide the large model to quickly generate the second material information required by the target object based on the reference example, thereby improving the working efficiency of the intelligent store building system, improving the accuracy of the generation of the second material information, and enhancing the satisfaction of the target object.

[0086] Step A2, input the prompt information into the target model to generate the second material information of the intelligent store.

[0087] The target model can be a large model. Different large models can be used for different second material information.

[0088] For example, in the case of second material information being text, a large language model can be used. The large language model refers to a specific type of large model that is specifically used for processing text data. Such models are natural language processing models based on neural networks and can be used to generate, understand and process text data. Some typical large language models have hundreds of billions of parameters and can generate high-quality text and can be used for various natural language processing tasks such as question answering, text generation, dialogue systems, etc.

[0089] For example, in the case of second material information being multimedia data such as pictures and videos, a multi-modal generation model can be used to build the second material information. Of course, the specific model structure, model parameter amount, etc. used in the embodiment of the present disclosure are not limited.

[0090] In the embodiment of the present disclosure, a small amount of first material information is used to generate second material information. The specific information content of the first material information is based on accumulated experience knowledge. The accumulated knowledge is converted into prompt information, and then the large model inference and prediction ability of the target model is relied on to automatically and efficiently generate the second material information. The generation of store materials can better understand the needs of the target object with less information, thereby providing personalized services for each target object, reducing the operation cost of the target object while improving the satisfaction of the target object. In addition, using the target model to process the prompt information realizes the automatic generation of the store materials, which has higher speed and efficiency. The target model quickly analyzes and infers when processing information with its powerful inference ability, improves the response speed and working efficiency of the system, and further improves the efficiency of the intelligent store generation.

[0091] In addition, in the embodiment of the present disclosure, a feedback mode is further added after the second material information is output, and the atomic generation capability of the second material information can be optimized through the feedback mode. For example, in the feedback mode, a modification operation on the second material information can be responded to, so as to achieve the effect of optimizing the generated second material information. In the embodiment of the present disclosure, the active expression form of the second material information is obtained through the active modification of the target object, and the specific execution process can include the following steps:

[0092] Step C1, output the second material information.

[0093] That is, the intelligent store building system can output the generated second material information to the target object for display.

[0094] Step C2, in response to a modification operation on the second material information, an update result of the second material information is obtained.

[0095] In the page where the second material information is displayed, a user operation entrance can be provided, so that the target object can modify the generated second material information.

[0096] Step C3, record the corresponding relationship between the update result of the first material information and the second material information.

[0097] The intelligent store building system can collect the generated second material information and the updated second material information through the dotting operation, so as to record the entire generation process of the second material information, so as to analyze the deficiency and optimize the atomic generation capability.

[0098] In addition to outputting the generated second material information to the target object for correction, in the feedback closed loop, in the embodiment of the present disclosure, the output second material information is fed back to the target object and the intelligent store building system. In the case that the output second material information does not meet the requirements of the target object, as described above, the target object can modify the generated second material information to obtain the second material information that meets the active expression of the target object. At the same time, the intelligent store building system will respond to the modification operation of the target object, record the corresponding relationship between the update result of the first material information and the second material information. In the case that the output second material information does not meet the requirements of the intelligent store building system on the store material, the intelligent store building system will re-generate the second material information to meet the requirements of the corresponding risk control strategy, and the intelligent store building system will record the corresponding relationship between the update result of the first material information and the second material information. Of course, the update result here can be the result of the active expression of the target object, or the result of the automatic correction of the system. In short, the update result of the second material information can be understood as the second material information finally adopted by the target object.

[0099] In the embodiments of the present disclosure, the second material information is modified by means of the feedback information of the target object, so that the entire process of generating the second material information forms a complete closed loop. By continuously optimizing the generation result of the second material information, it can be ensured that the generated second material information meets the requirements of the target object, and the corresponding relationship between the second material information after the modification operation and the first material information is recorded, which is very helpful for avoiding similar problems and correcting errors in the future, so as to complete the optimization of the corresponding atomic generation capability, reduce the repeated work in the later use of the intelligent store generation system, and improve the efficiency of intelligent store generation.

[0100] In addition, the recorded corresponding relationship can enable the intelligent store to serve as a reference store to guide other stores to generate corresponding second material information.

[0101] In implementation, for the atomic generation capability of the store name, an intelligent store in the same industry can be used as a reference store; for the atomic generation capability of the commodity label, the corresponding relationship of the same type of commodity of an intelligent store in the same commodity category can be used to guide the generation of the commodity label of other intelligent stores. For the atomic generation capability of the multimedia resource, an intelligent store with the same industry and commodity category can be used as a reference store to construct a corresponding guide example.

[0102] In the embodiments of the present disclosure, as described above, the first material information includes the first type of sub-information, the second type of sub-information and the third type of sub-information for different atomic generation capabilities.

[0103] The first type of sub-information includes at least one of the following: the enterprise name of the intelligent store, the industry information of the intelligent store;

[0104] The second type of sub-information includes at least one of the following: the industry information of the intelligent store, the commodity name, the commodity specification, the picture, and the text;

[0105] The third type of sub-information includes at least one of the following: the store name, the commodity category, the commodity label, and the original image.

[0106] The specific process of generating the corresponding second material information for each type of sub-information of the first material information will be described in detail below for different atomic generation capabilities:

[0107] 1) Atomic generation capability of store name

[0108] This atomic generation capability can automatically generate a suitable store name based on the first type of sub-information.

[0109] ​​​​​​​​​​In the embodiments of the present disclosure, based on a large amount of information collected in the early stage, such as enterprise name and industry information, comprehensive analysis and summary are carried out to obtain a paradigm for generating a store name. Among them, the analysis of the enterprise name can be to extract the main word of the enterprise name. Finally, the paradigm of the store name can be represented as: the main word of the company extracted from the enterprise name + aggregation of classified industry information + strategy (first risk control strategy (such as word expression limit), industry category) + LLM (Large Language Model) = store name.

[0110] Among them, in order to generate store names suitable for different industries, the embodiments of the present disclosure adopt a few-shot prompt composition scheme. In the case where the first material information includes a first type of sub-information, the industry category of the intelligent store is determined based on the first type of sub-information. It can be understood that a large amount of user merchant data is obtained by clustering analysis or manual annotation to obtain different industry categories. For example, as shown in Figure 3 , N data are classified by industry to obtain the naming characteristics of four different industries. The four categories include Figure 3 service enterprises, ordinary companies, operating departments, and law firms. The classification results are combined with the first risk control strategy to constrain the generation of store names.

[0111] Based on the first type of sub-information, the industry category, and the first risk control strategy required for generating the store name, a first sub-prompt information is constructed; the prompt information includes the first sub-prompt information. For example, as shown in Figure 3 , the first type of sub-information is enterprise name and industry information, and the first sub-prompt information is obtained by combining the industry category and the first risk control strategy. Among them, the first risk control strategy can be at least one of word limit, special symbol limit, or necessary information required for generating the store name. For example, based on the enterprise name and industry information input by the target object, a store name is designed. The industry information is the main range that the store may operate, and the first risk control strategy requires that the output content only contains the store name and does not contain any other information or explanation, and cannot contain special characters such as

[0112] For example, {enterprise name: AC Network Technology Co., Ltd. Industry information: legal consultation. Store name: AC legal consultation}. Among them, the enterprise name and industry information are the first material information input by the target object, and the store name is the output result of the atomic generation capability of the store name.

[0113] In the embodiments of the present disclosure, the experience is precipitated to obtain the key guidance information of the industry category into the prompt information, which is used to construct the first sub-prompt information. Based on the industry category, the model can predict the available store name according to the learned knowledge. Meanwhile, the prompt information is constructed based on the first risk control strategy, which improves the availability and effect of the generated store name.

[0114] In the embodiments of the present disclosure, the target model can be a large language model. In the implementation, the first sub-prompt information is input into the large language model, so that the large language model generates the second material information under the guidance of the first sub-prompt information. At this time, for the atomic generation ability of the store name, in the case that the first material information is the first type of sub-information, the second material information is the store name.

[0115] As shown in Figure 3 , the enterprise name, the industry information, the industry category, and the first risk control strategy required for generating the store name construct the first sub-prompt information. The first sub-prompt information is input into the large language model LLM (Large Language Model), and the store name is output.

[0116] In addition, as shown in Figure 3 , in order to improve the atomic generation ability of the store name of the target model, the generated store name is also fed back to the target object and the system, so that the entire process of generating the store name forms a complete closed loop. By feeding back the generated store name to the target object, it can be judged whether the generated store name meets the requirements of the target object. In the case that the generated store name does not meet the requirements of the target object, the target object can modify the generated store name. At the same time, the intelligent store building system records the corresponding relationship between the first type of sub-information and the modified store name, which serves as a reference for the next round of store name generation. In addition, by feeding back the generated store name to the intelligent store building system, it can be judged whether the generated store name meets the requirements of the first risk control strategy, such as whether the number of words exceeds the limit, whether there are special symbols, etc. In the case that the generated store name does not meet the requirements of the first risk control strategy, the intelligent store building system generates the store name again, and records the corresponding relationship between the first type of sub-information and the updated store name. Through such a closed loop feedback, the target model can be gradually improved, so that it can gradually learn and master the generation ability of the store name.

[0117] In the embodiments of the present disclosure, the second material information meeting the requirements of the target object is generated by using a large language model under the guidance of the first prompt information, which can improve the store name generation capability of the target model and further save the time and effort of the target object. Through the prompt information, the model can automatically generate the second material information, reducing the operation cost of the target object and improving the efficiency of intelligent store generation. In addition, according to the first risk control strategy in the prompt information, personalized second material information can be generated while meeting the requirements of intelligent store generation, thereby improving the satisfaction of the target object with the intelligent store generation system.

[0118] 2) Atomic generation capability of product label

[0119] The atomic generation capability can automatically generate suitable product labels based on the second type of sub-information.

[0120] As key information for product information disclosure, product labels affect the interest of consumers in products. More accurate and comprehensive label information can greatly benefit the investment effect of the target object.

[0121] In the embodiments of the present disclosure, based on various product data information collected in the early stage, including but not limited to industry information, product name, product picture text, and selling specifications, comprehensive analysis is performed to finally obtain a paradigm that can be used to generate product labels. In order to generate product labels that meet the characteristics of different industries, and to simplify user operations, the embodiments of the present disclosure adopt a few-shot prompt composition scheme. Finally, through big data analysis and summary, the paradigm of the product label can be expressed as: product name + industry information + selling specifications + image / text + strategy (product category, second risk control strategy) + LLM (large language model) = product label. Among them, the image and text can be used alternatively or jointly. The image / text here can be used to describe the relevant product.

[0122] In the case where the first material information includes the second type of sub-information, the product category is determined based on the second type of sub-information. It can be understood that, by means of clustering analysis or manual marking, different product categories are obtained from a large amount of product information. For example, as shown in Figure 4 , N product data are classified to obtain a plurality of product categories. The plurality of product categories can be as shown in Figure 4 , including life service, education and training, real estate and home, etc. The classification result is combined with the second risk control strategy to constrain the output of the product label.

[0123] Based on the second type of sub-information, the product category, and the second risk control strategy required for generating the product label, the second prompt information is constructed; the prompt information includes the second prompt information. For example, as shown in Figure 4As shown, the second type of sub-information includes industry information, commodity name, picture / text of the commodity, and selling specifications, and the second sub-prompt information is obtained in combination of the commodity category and the second risk control strategy. The picture of the commodity can also include text, and the text information contained in the picture can be identified by using an OCR technology in the process of processing the picture. The second risk control strategy can include a word limit, a special symbol limit, or necessary information required for generating a commodity label. The specific risk control strategy requirements can be consistent with or different from the first risk control strategy requirements for generating the store name, and the second risk control strategy can be set according to actual needs in implementation, and the present embodiment does not make a detailed description here.

[0124] In the present embodiment, the second sub-prompt information requiring a small amount of information is constructed by classifying the commodities, and the system can generate a targeted commodity label. The second risk control strategy included in the second sub-prompt information can enable the large model to regulate the generated commodity label based on the requirements of the second risk control strategy, so that the commodity label can meet the requirements of intelligent store building and improve the efficiency of intelligent store building.

[0125] Similarly, in the present embodiment, the second sub-prompt information can be input into a large language model (LLM) to enable the large language model to generate second material information under the guidance of the second sub-prompt information. In the case where the first material information is the second type of sub-information, the second material information is the commodity label.

[0126] As shown, Figure 4 The industry information, commodity name, picture and / or text, selling specifications, commodity category, and second risk control strategy required for generating a commodity label construct the second sub-prompt information, and the second sub-prompt information is input into a large language model, thereby generating a commodity label.

[0127] In addition, similarly as described above, as shown, Figure 4As shown, the generated product label is also fed back to the target object and the intelligent store building system, so that the entire process of generating the product label forms a complete closed loop. By feeding the generated product label to the target object, it can be determined whether the generated product label meets the requirements of the target object. In the case where the generated product label does not meet the requirements of the target object, the target object can modify the generated product label. At the same time, the intelligent store building system records the correspondence between the second type of sub-information and the modified product label, so as to facilitate the construction of a guide example for subsequent generation of product labels. In addition, by feeding the generated product label to the intelligent store building system, it can be determined whether the generated product label meets the requirements of the second risk control strategy, such as whether the number of words exceeds the limit, whether there are special symbols, etc. In the case where the generated product label does not meet the requirements of the second risk control strategy, the system re-generates the product label and records the correspondence between the second type of sub-information and the updated product label. Through such a closed loop feedback, the model can be gradually improved to gradually learn and master the generation ability of product labels.

[0128] It should be emphasized that the product label actively modified and adopted by the target object will eventually be recorded in the system, so as to serve as a correct guide example in the process of generating product labels for the next product of the same category, guiding the model to generate product labels for products of the same category.

[0129] In the embodiments of the present disclosure, the use of a large language model under the guidance of the second sub-prompt information to generate second material information that meets the requirements of the target object can save the time and effort of the target object. Using only a small amount of prompt information can prompt the large language model to automatically generate the second material information, reducing the requirement for the amount of information input by the user. Based on the product category and the second risk control strategy, the large language model can be guided to generate a usable product label, improving the efficiency of intelligent store building.

[0130] 3) Atomic generation capability of multimedia resources

[0131] This atomic generation capability is based on third type of sub-information to generate multimedia resources. Wherein, the multimedia resources can include pictures and / or videos. Especially in the design of landing page. After directing traffic to the website, the most important thing is to create conversion rate, only when the website visitor and the enterprise establish relationship, the traffic has the opportunity to be converted into customers. The landing page is created to efficiently achieve the conversion number. The landing page is a website page used to attract a specific target audience and guide them to take a specific action, so that the intelligent store can obtain potential customers. That is, the landing page focuses on conversion rate. Therefore, multiple different landing pages are generally purposefully made for specific audiences and scenarios.

[0132] As an important page for attracting traffic, the design of the header image in the landing page is very important. Based on the atom generation capability, the landing page that meets the requirements can be effectively generated.

[0133] In the process of generating an intelligent store, the material information provided by the target object is often incomplete or does not meet the requirements. Generally, the original picture information or original video information provided is a material resource, without marketing attributes and strong styles and colors, and the definition also leads to the target object unable to reach the threshold of entering the intelligent store building system.

[0134] In the process of generating an intelligent store, the material information provided by the target object is often incomplete or does not meet the requirements. Generally, the original picture information or original video information provided is a material resource, without marketing attributes and strong styles and colors, and the definition also leads to the target object unable to reach the threshold of entering the intelligent store building system.

[0135] In the embodiment of the present disclosure, based on the comprehensive analysis of the multimedia resource data collected in the early stage. Finally, through big data analysis, a paradigm for generating multimedia resources with marketing attributes is summarized. The final generated multimedia resource paradigm can be represented as: category prediction + product label generation + store name extraction + industry information + multimedia resource color -> multimedia resource generation (original multimedia resource + cropping + stretching + blur + style superposition) = final multimedia resource.

[0136] In the embodiment of the present disclosure, based on the comprehensive analysis of the multimedia resource data collected in the early stage. Finally, through big data analysis, a paradigm for generating multimedia resources with marketing attributes is summarized. The final generated multimedia resource paradigm can be represented as: category prediction + product label generation + store name extraction + industry information + multimedia resource color -> multimedia resource generation (original multimedia resource + cropping + stretching + blur + style superposition) = final multimedia resource. Figure 5As shown, N multimedia resource data are clustered based on their multimedia resources, resulting in four main categories: color, style, layout, and text. For different industries, color distributions and styles applicable to those industries can be summarized as templates based on these four categories. Each style has corresponding text formatting and layout requirements. For a target object, an applicable template can be obtained based on the industry category to which the target object belongs, thus obtaining the target color distribution and target style applicable to the target object. Subsequently, a third risk control strategy can be further combined to constrain the final output multimedia resources.

[0137] Based on the third type of sub-information, target style and target color distribution, as well as the third risk control strategy required to generate product labels, a third sub-prompt information is constructed; the prompt information includes the third sub-prompt information.

[0138] like Figure 5 As shown, the third type of sub-information includes the original multimedia resources provided by the target object, the store name, product tags, and the product category prediction results. The store name can be obtained through the atomic generation capability of the aforementioned store name, the product tags can be obtained through the atomic generation capability of the aforementioned product tags, and the category prediction can be obtained through a prediction model. The prediction model can employ deep learning models, ensemble learning models, rule-based models, etc., and this disclosure does not impose any restrictions on it.

[0139] The third type of sub-information, along with the target style, target color distribution, and third risk control strategy, constitutes the third sub-prompt information. An example of the target style is as follows: Figure 6 As shown, you can constrain the position, format, and expression of product tags on the page. You can also constrain the position and format of the store name on the page.

[0140] Examples of target color distributions are as follows Figure 7 As shown, red can be used for the color scheme of moving and hauling goods, blue for digital repair, and green for on-site installation.

[0141] Of course, the color distribution can be a single color or a complex color scheme, and this disclosure does not limit this. The color distribution can constrain the background color of the corresponding text or the tonal feel of the image content. In implementation, the color distribution can be determined according to actual needs.

[0142] In this embodiment of the disclosure, the target style and target color distribution are determined by the third type of sub-information input by the target object, thereby determining the style requirements and color requirements suitable for the smart store, which can improve the efficiency of smart store generation.

[0143] In addition, the third sub-prompt information is generated in combination with the enterprise main word, the commodity label and the category prediction, which can enrich the content of the multimedia resource, intuitively and elegantly display the information of the target object on the multimedia resource, and improve the satisfaction of the target object. Moreover, under the constraint of the third risk control strategy, the generated multimedia resource is more standardized, which is helpful for the novice to open a store.

[0144] In the embodiments of the present disclosure, the third sub-prompt information can be input into a generation model, so that the generation model generates second material information under the guidance of the third sub-prompt information. Here, the second material information is the multimedia resource.

[0145] As shown in Figure 5 , the store name, the commodity label, the category prediction, the target color distribution, the target style, the original multimedia resource input by the target object and the third risk control strategy required for generating the final multimedia resource constitute the third sub-prompt information. Then, the third sub-prompt information is input into a generation model to output the final multimedia resource. The generation model can be composed of an LLM model and a multi-modal generation model. The store name and the commodity label are obtained through the LLM model, the multi-modal generation model is used to obtain the category prediction result of the multimedia resource, and then the third sub-prompt information is constructed and input into the generation model with multi-modal information processing capability to generate the multimedia resource. As shown in Figure 5 , the third sub-prompt information contains information of two modalities, i.e., image information and text information. The third sub-prompt information is input into the generation model with multi-modal information processing capability for processing, so that the generation model combines the input text, style, color distribution and strategy requirements to finally output the multimedia resource information meeting the requirements of intelligent store opening.

[0146] In addition, similar to the previous two atomic generation capabilities, as shown in Figure 5 , the final generated multimedia resource is also fed back to the target object and the intelligent store opening system, so that the entire process of generating the multimedia resource forms a complete closed loop. By feeding back the final generated multimedia resource to the target object, it can be determined whether the final generated multimedia resource meets the requirements of the target object. In the case where the final generated multimedia resource does not meet the requirements of the target object, the target object can modify the final generated multimedia resource. At the same time, the intelligent store opening system records the correspondence between the third type of sub-information and the modified multimedia resource, which is used for subsequent generation of multimedia resources, and the actual adopted multimedia resource of the target object can be used to construct a guide example. In addition, by feeding back the final generated multimedia resource to the intelligent store opening system, it can be determined whether the final generated multimedia resource meets the requirements of the third risk control strategy. In the case where the generated multimedia resource does not meet the requirements of the third risk control strategy, the intelligent store opening system re-generates the multimedia resource and records the correspondence between the third type of sub-information and the updated multimedia resource.

[0147] In the third sub-prompting information, the generation model is used to generate multimedia information meeting the requirements, which can improve the efficiency of intelligent store building, is more friendly to users who cannot design multimedia resources, and can improve user experience.

[0148] As described above, the first material information is divided into three categories, and based on the three categories of information, the generation process of the intelligent store is divided into the generation of the store name, the generation of the product label, and the generation of the multimedia resource. In the first material information classification process, the generation process of the intelligent store is also classified, the atomic generation capability of the corresponding store building material is constructed by combining the demand of the store building material with the minimum amount of information, and finally, based on the atomic generation capability, the store building path can be shortened, the operation convenience and efficiency of the target object are improved, the generated material information is more complete, and the efficiency and accuracy of the intelligent store generation are improved.

[0149] The information processing method provided in the embodiment of the present disclosure can guide the target object to input a small amount of necessary information by prompting the target object on demand in the form of a dialogue flow. The generation method reduces the amount of information input by the target object, and integrates all information on the operation path of the target object into a single dialogue flow. In the single dialogue flow, the user does not need to wait for the audit result of the audit link, and can help the target object to submit the first material information in one-stop mode. The subsequent intelligent store building system will create an audit environment in parallel and send the audit on demand according to the submitted material type and the audit type. Therefore, the operation efficiency of the target object can be improved, the operation path of the target object can be shortened, the number of fields required to be input by the target object can be reduced, the store building threshold of the target object can be lowered, and the target object can achieve good store building effect with a small number of steps. In the embodiment of the present disclosure, the first material information can be obtained by using the information processing method provided in the embodiment of the present disclosure. Figure 8a and Figure 8b The difference between the existing process and the process of the present disclosure is intuitively expressed, and from the comparison of the two figures, it can be seen that the method steps provided by the present disclosure are fewer. Referring to Figure 8a and Figure 8b It can be seen that the number of fields required to be input by the target object is reduced, thereby reducing the amount of information manually input by the user. The operation cost and time required for generating the intelligent store are shorter, and the construction efficiency of the intelligent store is significantly improved.

[0150] In order to facilitate understanding of the general process of intelligent store building, the embodiment of the present disclosure combines Figures 9a-9e to describe the intelligent store building system.

[0151] The target object enters the intelligent store building module of the intelligent store building system. First, the target object uploads the business license to the intelligent store building system under the guidance of the intelligent store building system and confirms the uploaded business license. Then the intelligent store building system guides the target object to ask questions and confirms whether the target object has an offline store. In the intelligent store building system, the target object needs to input less additional information, and most of the information can be input by clicking the selection operation. After confirming that the target object does not have an offline store, the intelligent store building system guides the target object to input the keywords required for generating an intelligent store. From Figure 9a It can be seen from the above example that in the case of generating a Y education intelligent store, the target object inputs the first character "Y", and the intelligent store building system will recommend candidate keywords based on previous experience, so that the target object can obtain the required keyword information without complex operation.

[0152] Figure 9b The operation after the target object confirms the business keyword as Y education is shown. From Figure 9b It can be seen from the above example that when the target object inputs the business keyword, the intelligent store building system will generate the main category that the target object may need according to the business keyword confirmed by the target object, and then the target object selects the main category required by the target object in the category table given by the intelligent store building system. After the target object confirms the main category, the target object uploads the category qualification according to the guidance of the intelligent store building system. The category qualification can be 2 products / services, and the name and the introduction of the category qualification need to be improved based on the risk strategy.

[0153] Figure 9c The operation after uploading the category qualification is shown. From Figure 9c It can be seen from the above example that after the intelligent store building system confirms that it has received the category qualification of the target object, the target object needs to input the contact number to obtain the clues of the target object. After receiving the contact number of the target object, the intelligent store building system shows the platform agreement of the intelligent store building platform. In the case that the target object views the generation of the intelligent store and promises to comply with the platform agreement, the target object is requested to confirm the platform agreement. After confirming the platform agreement, the target object improves the information in the right column according to the guidance of the intelligent store building system. After improving all the necessary information, click the start generation in the lower right corner, and the system will start generating the page according to the material information provided by the target object.

[0154] From the above example, it can be seen that the target object does not need to fill in the material information according to the existing form one by one, and only needs to fill in part of the information through the guided dialogue, and the intelligent store building system will generate additional other store building materials.

[0155] The first material information and the generated second material information are combined to build an intelligent store. The required store building materials include the basic information of the target object store, such as enterprise name, industry information, etc. These information can be displayed on the generated intelligent store page to help consumers clearly understand the background and characteristics of the target object. Based on these information, the store name can be generated using the atomic generation capability of the store name. At the same time, the target object can also add and manage product information in the intelligent store building system, including product name, selling specifications, etc. These information can also be displayed on the intelligent store page to help consumers understand the detailed information of the product. In addition, these information can be used to build the second type of sub-information to generate product labels based on the atomic generation capability of the product label. The target object can also upload picture information in the intelligent store building system to help consumers intuitively understand the store image of the target object. The picture information can also build the third type of sub-information to generate multimedia resources of the store based on the atomic generation capability of the multimedia resources.

[0156] For example, taking a home decoration store as an example, Figure 9d The store homepage and marketing effect page built by the intelligent store building system according to the first material information input by the target object and the generated second material information are shown. Among them, Figure 9d The store homepage view and marketing page view generated by the intelligent store building system according to the information of the target object are shown in 9d. As can be seen from 9d, the store homepage view contains the store name, and the marketing page view contains the industry information and product labels, which are displayed in the form of text and pictures. The design of the entire page is intuitive and exquisite. The target object can select the page view that needs to be adopted according to its own needs. Assuming that the target object selects the store homepage view and marketing page 1 and marketing page 2 views to be adopted, and then enters the interface shown in Figure 9e After the target object confirms that the generated page view is correct, it clicks on "Local store promotion" to complete the generation of the intelligent store of the target object.

[0157] In the embodiments of the present disclosure, the ability to automatically generate information can be improved by continuously improving the atomic capabilities of the intelligent store building system. By building the following process: adoption of dotting -> data presentation -> low-quality alarm -> effect analysis -> prompt dynamic second-level update, a complete closed loop from atomic capability online to effect improvement can be achieved. For example, in the entire intelligent store building system, corresponding processing records can be collected by dotting, and abnormal data can be collected, aggregated and displayed, which can provide data support for further optimizing atomic generation capabilities. When the data generated by the atomic generation capability exceeds a certain threshold and is not adopted by the target object, or does not meet the risk control strategy, an alarm can be given to optimize the atomic generation capability. In the case of active modification of the target object and active expression of information content, the actively expressed information can be used as a guide example for the next round of information generation to achieve dynamic second-level update. Finally, the embodiments of the present disclosure provide a plurality of atomic generation capabilities to generate store building materials. When implemented, the atomic generation capabilities can be assembled according to the needs of the implementation scenario. In the embodiments of the present disclosure, based on the atomic AI generation capability, the local business itself is combined to complete the module assembly of AI generation capability and business empowerment.

[0158] For example, as shown in Figure 10 , the intelligent store building system can complete the abstraction of 4 core main businesses and 20+ business atomic services. The 4 core main businesses mainly include store atomic information, merchant atomic information, decoration atomic information, and commodity atomic information. By building AIGC conversation stringing, idempotent retry, and ID preposition mechanisms, the system stability is guaranteed while supporting the completeness of data and meeting the efficiency of service operation; among them, the ID preposition is the pre-generated core material ID, i.e. the conversation ID described above, which guarantees the continuity of the subsequent process of the upstream system and reduces the time consumption of the overall scheduling. Each module can be divided into independent sub-processes. Some sub-processes can collect and aggregate different information groups and hand them over to another sub-process to complete the information review work in parallel; the idempotent retry is to perform idempotent execution after capturing the exception through the retry task at any step, which guarantees the completeness of AI store and commodity generation.

[0159] In summary, in the related art, the degree of active participation of the target object basically determines the effect of the final page and the conversion effect expected by the customer. In this process, information such as business license identification information, name generation, and tag generation usually needs to be actively provided by the customer. However, the target object can only obtain a page style with an initialization effect, which increases the operation cost of the customer. The method provided by the embodiments of the present disclosure mainly aims to improve the automatic generation capability of the system and simplify the process, thereby reducing the operation and understanding cost of the target object, and generating store building material information such as stores and commodities with both effect and efficiency in a systematic way.

[0160] Based on the same technical concept, the embodiments of the present disclosure further provide an information processing apparatus 1100, as shown in the accompanying drawings, comprising: Figure 11

[0161] The first obtaining module 1101 is configured to obtain first material information of the intelligent store.

[0162] The generating module 1102 is configured to generate second material information of the intelligent store based on the first material information.

[0163] The first material information and the second material information construct the intelligent store.

[0164] In some embodiments, the generating module comprises:

[0165] The first generating sub-module is configured to generate prompt information based on the first material information.

[0166] The second generating sub-module is configured to input the prompt information into a target model to generate the second material information of the intelligent store.

[0167] In some embodiments, in a case where the first material information is a first type of sub-information required for generating a store name, the second material information is the store name.

[0168] In a case where the first material information is a second type of sub-information required for generating a product label, the second material information is the product label.

[0169] In a case where the first material information is a third type of sub-information required for generating a multimedia resource, the second material information is the multimedia resource.

[0170] In some embodiments, the first generating sub-module is specifically configured to:

[0171] In a case where the first material information comprises the first type of sub-information, determine an industry category of the intelligent store based on the first type of sub-information.

[0172] Construct a first sub-prompt information based on the first type of sub-information, the industry category, and a first risk control strategy required for generating the store name; and the prompt information comprises the first sub-prompt information.

[0173] In some embodiments, the first generating sub-module is specifically configured to:

[0174] In a case where the first material information comprises the second type of sub-information, determine a product category based on the second type of sub-information.

[0175] Construct a second sub-prompt information based on the second type of sub-information, the product category, and a second risk control strategy required for generating the product label; and the prompt information comprises the second sub-prompt information.

[0176] ​In some embodiments, the first generation submodule is specifically configured to:

[0177] In a case where the first material information includes the third type of sub-information, determining the target style and the target color distribution of the applicable smart store based on the third type of sub-information;

[0178] Based on the third type of sub-information, the target style, the target color distribution, and a third risk control strategy required for generating the product label, a third sub-prompt information is constructed; the prompt information includes the third sub-prompt information.

[0179] In some embodiments, the second generation submodule is specifically configured to:

[0180] The first sub-prompt information is input into the large language model, so that the large language model generates the second material information under the guidance of the first sub-prompt information.

[0181] In some embodiments, the second generation submodule is specifically configured to:

[0182] The second sub-prompt information is input into the large language model, so that the large language model generates the second material information under the guidance of the second sub-prompt information.

[0183] In some embodiments, the second generation submodule is specifically configured to:

[0184] The third sub-prompt information is input into the generation model, so that the generation model generates the second material information under the guidance of the third sub-prompt information.

[0185] In some embodiments, further comprising:

[0186] The output module is configured to output the second material information;

[0187] The second acquisition module is configured to acquire an update result of the second material information in response to a modification operation on the second material information;

[0188] The recording module is configured to record a corresponding relationship between the update result of the first material information and the second material information.

[0189] In some embodiments, the first generation submodule is specifically configured to:

[0190] The reference store of the smart store is found;

[0191] Based on the corresponding relationship between the first material information of the reference store and the update result of the second material information of the reference store, a guidance example is constructed;

[0192] Based on the first material information of the smart store and the guidance example, the prompt information is constructed.

[0193] In some embodiments, the first acquisition module comprises:

[0194] The acquisition sub-module is configured to acquire first material information of the intelligent store expected by the target object by interacting with the target object.

[0195] In some embodiments, the method further comprises:

[0196] The identification module is configured to identify, in the process of interacting with the target object, an information group belonging to the target quality inspection link based on the session mark.

[0197] The quality inspection module is configured to input the information group into the target quality inspection link to complete the quality inspection operation.

[0198] In some embodiments, the first type of sub-information includes at least one of the following: an enterprise name of the intelligent store, industry information of the intelligent store.

[0199] The second type of sub-information includes at least one of the following: industry information of the intelligent store, a commodity name, a commodity specification, a picture, and text.

[0200] The third type of sub-information includes at least one of the following: a store name, a commodity category, a commodity label, and an original image.

[0201] The specific functions and examples of the modules and sub-modules of the apparatus of the embodiments of the present disclosure are described in the related description of the corresponding steps in the above method embodiments, which will not be described here.

[0202] According to the embodiments of the present disclosure, the present disclosure further provides an electronic device, a readable storage medium and a computer program product.

[0203] Figure 12 A schematic block diagram of an example electronic device 1200 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not intended to limit the implementations of the present disclosure described and / or claimed in this document.

[0204] As Figure 12As shown, the apparatus 1200 includes a computing unit 1201 that can perform various appropriate actions and processes in accordance with a computer program stored in a read-only memory (ROM) 1202 or a computer program loaded into a random access memory (RAM) 1203 from a storage unit 1208. Various programs and data required for the operation of the apparatus 1200 can also be stored in the RAM 1203. The computing unit 1201, the ROM 1202, and the RAM 1203 are connected to each other through a bus 1204. An input / output (I / O) interface 1205 is also connected to the bus 1204.

[0205] A plurality of components in the apparatus 1200 are connected to the I / O interface 1205, including an input unit 1206 such as a keyboard, a mouse, etc., an output unit 1207 such as various types of displays, speakers, etc., a storage unit 1208 such as a magnetic disk, an optical disk, etc., and a communication unit 1209 such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1209 allows the apparatus 1200 to exchange information / data with other apparatuses through a computer network such as the Internet and / or various telecommunication networks.

[0206] The computing unit 1201 can be various general and / or special purpose processing components having processing and computing capabilities. Some examples of the computing unit 1201 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 1201 performs various methods and processes described above, such as the information processing method. For example, in some embodiments, the information processing method can be implemented as a computer software program that is tangibly embodied in a machine-readable medium, such as the storage unit 1208. In some embodiments, part or all of the computer program can be loaded and / or installed on the apparatus 1200 via the ROM 1202 and / or the communication unit 1209. When the computer program is loaded into the RAM 1203 and executed by the computing unit 1201, one or more steps of the information processing method described above can be performed. Alternatively, in other embodiments, the computing unit 1201 can be configured to perform the information processing method by any other appropriate means, such as by means of firmware.

[0207] The various embodiments of the systems and techniques described above can be implemented in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a complex programmable logic device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0208] Program code for carrying out methods of the present disclosure can be written in any combination of one or more programming languages. The program code can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the program code, when executed by the processor or controller, produces the functions / operations specified in the flowcharts and / or the block diagrams. The program code can be executed entirely on a machine, partially on a machine, partially on a machine as a stand-alone software package, or entirely on a remote machine or server.

[0209] In the context of the present disclosure, a machine-readable medium can be a tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include but is not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk drives, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0210] To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0211] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0212] The computer system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0213] According to an embodiment of the present disclosure, the electronic device can be integrally integrated with the communication component, the display screen, and the information collection device, or can be separately provided from the communication component, the display screen, and the information collection device.

[0214] It should be understood that various forms of flow shown above can be used, with steps reordered, added, or removed. For example, the steps described in the present disclosure can be executed in parallel, in sequence, or in a different order, as long as the desired results of the technical solutions disclosed in the present disclosure can be achieved, and the present disclosure is not limited herein.

[0215] The foregoing detailed description has set forth various embodiments of the devices and / or methods via the use of specific terminology. However, embodiments thereof can be practiced with the exact description not being presented in detail. The term "device" as used herein should be interpreted to include devices and / or methods. The embodiments disclosed herein are illustrative of the principles of the present disclosure. Numerous modifications and adaptations will be readily apparent to those skilled in the art without departing from the spirit and scope of the present disclosure.

Claims

1. An information processing method, comprising: obtaining first material information of a smart store; generating prompt information based on the first material information; inputting the prompt information into a target model to generate second material information of the smart store; the second material information comprising a store name and a product label; in a case where the first material information comprises third type sub-information, determining a template applicable to the smart store based on an industry category and the third type sub-information; the template comprising a target style and a target color distribution; the template being obtained based on clustering analysis on a plurality of multimedia resources; the third type sub-information comprising: the store name, a product category, the product label, original multimedia resources; the target style, used to constrain a position and a format of the product label in a generated page, and an expression manner; and to constrain a position and a format of the store name in the page; the industry category being generated based on first type sub-information required for generating the store name; a naming feature of the industry category being used to constrain the target model to generate the store name; the first type sub-information comprising at least one of: an enterprise name of the smart store, and industry information of the smart store; constructing third sub-prompt information based on the third type sub-information, the target style and the target color distribution, and a third risk control strategy required for generating multimedia resources; the prompt information comprising the third sub-prompt information; generating multimedia resources in the second material information based on the third sub-prompt information; the first material information and the second material information being used to construct the smart store.

2. The method of claim 1, in a case where the first material information is first type sub-information required for generating a store name, the second material information is the store name; in a case where the first material information is second type sub-information required for generating a product label, the second material information is the product label.

3. The method of claim 2, wherein, The generating prompt information based on the first material information comprises: in a case where the first material information comprises the first type sub-information, determining an industry category of the smart store based on the first type sub-information; constructing first sub-prompt information based on the first type sub-information, the industry category, and a first risk control strategy required for generating the store name; the prompt information comprising the first sub-prompt information.

4. The method of claim 2, wherein, The generating prompt information based on the first material information comprises: in a case where the first material information comprises the second type sub-information, determining a product category of the smart store based on the second type sub-information; constructing second sub-prompt information based on the second type sub-information, the product category, and a second risk control strategy required for generating the product label; the prompt information comprising the second sub-prompt information.

5. The method of claim 3, wherein, The inputting the prompt information into a target model to generate second material information of the smart store comprises: inputting the first sub-prompt information into a large language model, so that the large language model generates the second material information under the guidance of the first sub-prompt information.

6. The method of claim 4, wherein, The prompt information is input into a target model to generate second material information of the intelligent store. The second sub-prompt information is input into a large language model to enable the large language model to generate the second material information under the guidance of the second sub-prompt information.

7. The method of claim 1, wherein, The target model is a generation model for the third sub-prompt information.

8. The method of any one of claims 1-7, further comprising: outputting the second material information; in response to a modification operation on the second material information, obtaining an update result of the second material information; recording a correspondence relationship between the first material information and the update result of the second material information.

9. The method of any one of claims 1-7, wherein, The prompt information is generated based on the first material information, comprising: finding a reference store of the intelligent store; based on the correspondence relationship between the first material information of the reference store and the update result of the second material information of the reference store, constructing a guidance example; based on the first material information of the intelligent store and the guidance example, constructing the prompt information.

10. The method of any one of claims 1-7, wherein, The first material information of the intelligent store is obtained, comprising: interacting with a target object to obtain the first material information of the intelligent store expected by the target object.

11. The method of claim 10, further comprising: during the interaction with the target object, identifying an information group belonging to a target quality inspection link based on a conversation mark; inputting the information group into the target quality inspection link to complete a quality inspection operation.

12. The method of any one of claims 2-7, wherein the second type of sub-information comprises at least one of the following: industry information of the intelligent store, a product name, a product specification, a picture, and a text.

13. An information processing apparatus, comprising: a first obtaining module configured to obtain first material information of an intelligent store; a first generating module configured to generate prompt information based on the first material information; a second generating module configured to input the prompt information into a target model to generate second material information of the intelligent store; the second material information comprising a store name and a product label; when the first material information comprises third type of sub-information, the first generating module is specifically configured to determine a template applicable to the intelligent store based on an industry category and the third type of sub-information; the template comprises a target style and a target color distribution; the template is obtained based on clustering analysis of a plurality of multimedia resources; the third type of sub-information comprises the store name, a product category, the product label, and original multimedia resources. The target style is used to constrain the position and format of the product label in the generated page, and to constrain the position and format of the store name in the page; the third sub-prompt information is constructed based on the third type of sub-information, the target style, and a target color distribution, and a third risk control strategy required for generating a multimedia resource; the prompt information includes the third sub-prompt information; the industry category is generated based on the first type of sub-information required for generating the store name; the naming characteristics of the industry category are used to constrain the target model to generate the store name; the first type of sub-information includes at least one of the following: the enterprise name of the intelligent store, and the industry information of the intelligent store; The second generation module is specifically configured to generate the multimedia resource in the second material information based on the third sub-prompt information. The first material information and the second material information are used to construct the intelligent store.

14. The apparatus according to claim 13, wherein, in a case where the first material information is the first type of sub-information required for generating a store name, the second material information is the store name; and in a case where the first material information is the second type of sub-information required for generating a product label, the second material information is the product label. The first generation module is specifically configured to:

15. The apparatus of claim 14, wherein, in a case where the first material information includes the first type of sub-information, determine an industry category of the intelligent store based on the first type of sub-information; construct first sub-prompt information based on the first type of sub-information, the industry category, and a first risk control strategy required for generating the store name; and the prompt information includes the first sub-prompt information. The first generation module is specifically configured to:

16. The apparatus of claim 14, wherein, in a case where the first material information includes the second type of sub-information, determine a product category of the intelligent store based on the second type of sub-information; construct second sub-prompt information based on the second type of sub-information, the product category, and a second risk control strategy required for generating the product label; and the prompt information includes the second sub-prompt information. The second generation module is specifically configured to:

17. The apparatus of claim 15, wherein, input the first sub-prompt information into a large language model, so that the large language model generates the second material information under the guidance of the first sub-prompt information. The second generation module is specifically configured to:

18. The apparatus of claim 16, wherein, input the second sub-prompt information into a large language model, so that the large language model generates the second material information under the guidance of the second sub-prompt information. The target model is a generation model for the third sub-prompt information.

19. The apparatus of claim 13, wherein, 20. The apparatus according to any one of claims 13-19, further comprising: an output module configured to output the second material information; a second acquisition module configured to, in response to a modification operation on the second material information, acquire an update result of the second material information; a recording module configured to record a correspondence relationship between the update results of the first material information and the second material information. The first generation module is specifically configured to:

21. The apparatus of any of claims 13-19, wherein, find a reference store of the intelligent store; ​ construct a guide example based on a correspondence relationship between the first material information of the reference store and an update result of the second material information of the reference store; construct the prompt information based on the first material information of the intelligent store and the guide example.

22. The apparatus of any one of claims 13-19, wherein, The first acquisition module comprises: An acquisition sub-module configured to interact with a target object and acquire the first material information of the intelligent store expected by the target object.

23. The apparatus of claim 22, further comprising: an identification module configured to identify, based on a session mark, an information group belonging to a target quality inspection link during the interaction with the target object; a quality inspection module configured to input the information group into the target quality inspection link to complete a quality inspection operation.

24. The apparatus of any one of claims 14-19, wherein the second type of sub-information comprises at least one of the following: industry information of the intelligent store, a commodity name, a commodity specification, a picture, and a text.

25. An electronic device comprising: at least one processor; and a memory connected in communication with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-12.

26. A non-transitory computer readable storage medium having stored thereon computer instructions, wherein, The computer instructions are used to enable the computer to perform the method of any one of claims 1-12.

27. A computer program product comprising a computer program which, when executed by a processor, implements the method of any one of claims 1-12.

Images