Information processing method and apparatus

Abstract

Provided in the embodiments of the present disclosure are an information processing method and apparatus. The information processing method comprises: acquiring original commodity information of commodities of a target category in a target platform, and pre-processing the original commodity information into target commodity information corresponding to a preset information structure of the target platform; on the basis of the target commodity information, aggregating the commodities of the target category, so as to obtain a target commodity set, and calculating trend proportions respectively corresponding to target commodities included in the target commodity set and a plurality of pieces of commodity trend information; according to the trend proportions, selecting target commodity trend information from among the plurality of pieces of commodity trend information; using a large language model to process the target commodity trend information according to a preset commodity recruitment prompt word, so as to obtain commodity recruitment theme information, wherein the commodity recruitment theme information is used for recommending a commodity releasing strategy to a target merchant.

Description

Information processing methods and devices

[0001] This disclosure claims priority to Chinese Patent Application No. 202510389289.9, filed on March 28, 2025, with the title “Information Processing Method and Apparatus”, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This disclosure relates to the field of information processing technology, and in particular to information processing methods and apparatus. Background Technology

[0003] With the development of computer and internet technology, online shopping has become an important way for users to purchase goods. To provide users with a high-quality online shopping experience, platforms typically analyze transaction data to identify the core selling points of products. By recommending these core selling points to merchants, they can encourage merchants to release products that meet user purchasing needs, improving both the user shopping experience and the merchants' sales. Currently, when platforms analyze product core selling points, they mostly rely on manual browsing of site data, analyzing upcoming trends through rankings and other methods, then identifying the core selling points of trending products, and finally sending targeted recruitment information to guide merchants to release new products. While this method can achieve the desired results, it requires human resources and is susceptible to subjective human influence, leading to a lack of unified standards for trend analysis and inaccurate trend identification. This results in low efficiency and inaccuracy in generating recruitment information. Therefore, an effective solution is urgently needed to address these problems. Summary of the Invention

[0004] In view of this, embodiments of this disclosure provide an information processing method. One or more embodiments of this disclosure also relate to an information processing apparatus, a computing device, a computer-readable storage medium, and a computer program product, to address the technical deficiencies existing in the prior art.

[0005] According to a first aspect of the present disclosure, an information processing method is provided, comprising:

[0006] Obtain the original product information of the target category products in the target platform, and preprocess the original product information into target product information corresponding to the preset information structure of the target platform;

[0007] Aggregate the target category products based on the target product information to obtain a target product set, and calculate the trend percentage of the target products and multiple product trend information contained in the target product set respectively;

[0008] Filter the target product trend information from the multiple product trend information according to the trend proportion;

[0009] The target product trend information is processed using a large language model according to preset product suggestion words to obtain product theme information, wherein the product theme information is used to recommend product release strategies to target merchants.

[0010] According to a second aspect of the embodiments of this disclosure, another information processing method is provided, including:

[0011] Obtain the original product information of the target category products in the target platform, and preprocess the original product information into target product information corresponding to the preset information structure of the target platform;

[0012] Aggregate the target category products based on the target product information to obtain a target product set, and calculate the trend percentage of the target products and multiple product trend information contained in the target product set respectively;

[0013] According to the trend proportion, target product trend information is filtered from multiple product trend information, and the target product trend information is processed using a large language model according to preset product prompt words to obtain product theme information;

[0014] Select target merchants that match the product theme information from among the multiple merchants associated with the target platform, and recommend the product theme information to the target merchants.

[0015] According to a third aspect of the present disclosure, an information processing apparatus is provided, comprising:

[0016] The acquisition module is configured to acquire the original product information of the target category products in the target platform, and preprocess the original product information into target product information corresponding to the preset information structure of the target platform;

[0017] The aggregation module is configured to aggregate the target category products based on the target product information to obtain a target product set, and to calculate the trend percentage of the target products and multiple product trend information contained in the target product set.

[0018] The filtering module is configured to filter target product trend information from the plurality of product trend information according to the trend proportion;

[0019] The processing module is configured to use a large language model to process the target product trend information according to preset product prompts to obtain product theme information, wherein the product theme information is used to recommend product release strategies to target merchants.

[0020] According to a fourth aspect of the present disclosure, another information processing apparatus is provided, comprising:

[0021] The information acquisition module is configured to acquire the original product information of the target category products in the target platform, and preprocess the original product information into target product information corresponding to the preset information structure of the target platform;

[0022] The product aggregation module is configured to aggregate the target category products based on the target product information to obtain a target product set, and calculate the trend percentage of the target products and multiple product trend information contained in the target product set respectively;

[0023] The information filtering module is configured to filter target product trend information from multiple product trend information according to the trend proportion, and process the target product trend information using a large language model according to preset product prompt words to obtain product theme information;

[0024] The merchant matching module is configured to filter target merchants that match the product theme information from multiple merchants associated with the target platform, and recommend the product theme information to the target merchants.

[0025] According to a fifth aspect of the present disclosure, a computing device is provided, comprising:

[0026] Memory and processor;

[0027] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the above-described information processing method.

[0028] According to a sixth aspect of the present disclosure, a computer-readable storage medium is provided that stores computer-executable instructions that, when executed by a processor, implement the steps of the information processing method described above.

[0029] According to a seventh aspect of the present disclosure, a computer program product is provided, including a computer program or instructions that, when executed by a processor, implement the steps of the information processing method described above.

[0030] The information processing method provided in this embodiment, in order to improve the efficiency of product information generation while ensuring accuracy, can first preprocess the original product information of the target category products in the target platform into target product information with the corresponding preset information structure of the target platform after obtaining the original product information of the target category products, thereby realizing the unification of the information structure of multi-source product information. Then, the target category products can be aggregated according to the target product information to obtain a target product set, so that products with the same attributes are grouped into the same set. On this basis, the trend proportions of the target products contained in the target product set and the trend information of multiple products can be calculated respectively, and the target product trend information can be filtered from the multiple product trend information according to the trend proportions; thus realizing the analysis of the proportions of existing products relative to the preset product trend information to sort out the product selling point trends. Afterwards, the large language model can be used to process the trend information of target products according to the preset product recommendation prompts to obtain product recommendation theme information. This enables the recommendation of product release strategies to target merchants based on the product recommendation theme information. By analyzing the target category products within the platform, product recommendation theme information can be automatically generated, which not only ensures the accuracy of information generation but also improves the efficiency of product recommendation on the platform. As a result, merchants can quickly complete targeted product recommendation based on the generated product recommendation theme information, thereby boosting the platform's transaction volume and improving the user's product purchasing experience. Attached Figure Description

[0031] Figure 1 is a schematic diagram of an information processing method provided in an embodiment of this disclosure;

[0032] Figure 2 is a flowchart of an information processing method provided in an embodiment of this disclosure;

[0033] Figure 3 is a flowchart of another information processing method provided in an embodiment of this disclosure;

[0034] Figure 4 is a flowchart of an information processing method provided in an embodiment of this disclosure;

[0035] Figure 5 is a schematic diagram of the structure of an information processing device provided in an embodiment of the present disclosure;

[0036] Figure 6 is a schematic diagram of the structure of another information processing device provided in an embodiment of the present disclosure;

[0037] Figure 7 is a structural block diagram of a computing device provided in an embodiment of this disclosure. Detailed Implementation

[0038] Numerous specific details are set forth in the following description to provide a full understanding of this disclosure. However, this disclosure can be implemented in many other ways than those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this disclosure. Therefore, this disclosure is not limited to the specific implementations disclosed below.

[0039] The terminology used in one or more embodiments of this disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of this disclosure. The singular forms “a,” “the,” and “the” as used in one or more embodiments of this disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this disclosure refers to and includes any or all possible combinations of one or more associated listed items.

[0040] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this disclosure, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this disclosure, and similarly, second may also be referred to as first. Depending on the context, the word “if” as used herein may be interpreted as “when”, “in response to a determination”, or “when…”.

[0041] Furthermore, it should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in one or more embodiments of this disclosure are all information and data authorized by the user or fully authorized by all parties. Moreover, the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0042] In one or more embodiments of this disclosure, a large model refers to a deep learning model with a large number of model parameters, typically containing hundreds of millions, tens of billions, hundreds of billions, trillions, or even tens of trillions of model parameters. A large model can also be called a foundation model. It is pre-trained using large-scale unlabeled corpora to produce a pre-trained model with hundreds of millions of parameters. Such models can adapt to a wide range of downstream tasks and have good generalization ability. Examples include Large Language Models (LLMs) and multi-modal pre-training models.

[0043] In practical applications, large models only require a small number of samples to fine-tune the pre-trained model before they can be applied to different tasks. Large models can be widely used in fields such as Natural Language Processing (NLP) and Computer Vision. Specifically, they can be applied to computer vision tasks such as Visual Question Answering (VQA), Image Captioning (IC), and Image Generation, as well as NLP tasks such as text-based sentiment classification, text summarization, and machine translation. The main application scenarios for large models include digital assistants, intelligent robots, search, online education, office software, e-commerce, and intelligent design.

[0044] First, the terms and concepts involved in one or more embodiments of this disclosure will be explained.

[0045] Targeted product recruitment: This refers to the platform proactively guiding merchants to publish specific types of products based on market trends or specific goals, such as those targeting upcoming holidays or seasonal demands.

[0046] Product listing theme: This is a collection of information that includes the product's core selling points, images, and a brief description, designed to attract merchants to list their products.

[0047] Attribute mapping: Mapping product attributes from different platforms or data sources to a unified attribute system of the platform to ensure data consistency and comparability.

[0048] Category prediction: Based on information such as the product's name, description, and images, predict the category it belongs to on the platform.

[0049] Large language models are artificial intelligence models trained on massive amounts of text data. Their core idea is to capture the complexity and diversity of language through unsupervised learning. These models process and understand enormous amounts of language data, typically ranging from billions to hundreds of billions. They are able to learn and understand the grammatical, semantic, and contextual information in text, thereby generating coherent and logical natural language text.

[0050] This disclosure provides an information processing method, and also relates to an information processing apparatus, a computing device, a computer-readable storage medium, and a computer program product, which will be described in detail in the following embodiments.

[0051] Referring to the schematic diagram in Figure 1, the information processing method provided in this embodiment, in order to improve the efficiency of product information generation while ensuring accuracy, can first preprocess the original product information of the target category products in the target platform into target product information with the corresponding preset information structure of the target platform after obtaining the original product information of the target category products, thereby realizing the unification of the information structure of multi-source product information. Then, the target category products can be aggregated according to the target product information to obtain a target product set, so that products with the same attributes form the same set. On this basis, the trend proportions of the target products contained in the target product set and the trend information of multiple products can be calculated respectively, and the target product trend information can be filtered from the multiple product trend information according to the trend proportions; thus realizing the analysis of the proportions of existing products relative to the preset product trend information to sort out the product selling point trends. Afterwards, the large language model can be used to process the trend information of target products according to the preset product recommendation prompts to obtain product recommendation theme information. This enables the recommendation of product release strategies to target merchants based on the product recommendation theme information. By analyzing the target category products within the platform, product recommendation theme information can be automatically generated, which not only ensures the accuracy of information generation but also improves the efficiency of product recommendation on the platform. As a result, merchants can quickly complete targeted product recommendation based on the generated product recommendation theme information, thereby boosting the platform's transaction volume and improving the user's product purchasing experience.

[0052] Referring to Figure 2, Figure 2 shows a flowchart of an information processing method provided according to an embodiment of the present disclosure, which specifically includes the following steps.

[0053] Step S202: Obtain the original product information of the target category products in the target platform, and preprocess the original product information into target product information corresponding to the preset information structure of the target platform.

[0054] The information processing method provided in this embodiment is applied to analyze target category products on a target platform, thereby predicting future sales trends of the products and generating product recruitment theme information corresponding to the trend information. This product recruitment theme information can then be published or recommended to merchants, enabling them to publish products according to the product recruitment theme information, thereby reaching more users, meeting users' shopping needs, and increasing merchants' product sales.

[0055] Specifically, the target platform refers to a platform that supports merchants selling goods. Correspondingly, the target category of goods refers to the collection of all categories of goods sold on the platform, or a collection of some categories of goods. All categories of goods refer to all types of goods sold on the platform, while some categories refer to a specific set of goods. In practical applications, the selection of some categories of goods can be set according to needs, such as clothing, computers, daily necessities, etc. This embodiment does not impose any limitations. Original product information refers to the product information corresponding to each product, including but not limited to product images, product titles, product attributes, product sales, product reviews, product descriptions, and product search volume, used for subsequent trend analysis. Correspondingly, preprocessing refers to data cleaning and structural unification of the original product information corresponding to each product, so that the original product information from different sources can have a unified information structure for easy subsequent use. Correspondingly, the preset information structure refers to the standard information structure set by the target platform, used to unify the use of original product information from different sources. Correspondingly, target product information refers to the product information obtained after structural unification processing of original product information from different sources.

[0056] Therefore, to improve the efficiency of product information generation while ensuring accuracy, after obtaining the original product information of the target category products on the target platform, the original product information can be preprocessed into target product information with the corresponding preset information structure of the target platform, thereby achieving a unified information structure for multi-source product information. Then, target category products can be aggregated based on the target product information to obtain a target product set, ensuring that products with the same attributes form the same set. Based on this, the trend proportions of the target products included in the target product set and multiple product trend information can be calculated, and the target product trend information can be filtered from multiple product trend information according to the trend proportions. This allows for the analysis of the proportions of existing products relative to preset product trend information to identify product selling point trends. Afterwards, the large language model can be used to process the trend information of target products according to the preset product recommendation prompts to obtain product recommendation theme information. This enables the recommendation of product release strategies to target merchants based on the product recommendation theme information. By analyzing the target category products within the platform, product recommendation theme information can be automatically generated, which not only ensures the accuracy of information generation but also improves the efficiency of product recommendation on the platform. As a result, merchants can quickly complete targeted product recommendation based on the generated product recommendation theme information, thereby boosting the platform's transaction volume and improving the user's product purchasing experience.

[0057] Furthermore, in order to process the product information of the target category into a unified information structure, a multimodal large model can be used. In this embodiment, the specific implementation is as follows:

[0058] The system acquires product details, product attributes, and product transaction information for a target category of goods on a target platform, and generates original product information based on these information. It then inputs the original product information into a multimodal big data model, updates it using standardized prompts corresponding to the target platform's preset information structure within the model, and obtains intermediate product information. Finally, it processes the intermediate product information using the multimodal big data model to obtain the target product information.

[0059] Specifically, product details refer to the description of any given product, such as its purpose, usage instructions, pairing options, and images. Correspondingly, product attribute information refers to the attribute description of any given product, such as its specifications, material, and dimensions. Similarly, product transaction information refers to the transaction description of any given product, such as its price, sales volume, reviews, and search volume. A multimodal large-scale model refers to a large language model capable of inputting information from different modalities and organizing it according to standardized prompts into an information structure that conforms to the target platform's preset structure. Standardized prompts are prompts constructed according to the target platform's preset information structure, used to ensure that after inputting original product information from different sources, the model outputs product information with the same structure for subsequent use.

[0060] Based on this, in order to improve the accuracy of subsequent product theme information construction, we can first obtain the product details, product attributes, and product transaction information of the target category products on the target platform. At this point, raw product information can be generated based on the product details, product attributes, and product transaction information. On this basis, the raw product information corresponding to each product can be input into the multimodal big data model. The raw product information is updated by the standardized prompts corresponding to the preset information structure of the target platform in the multimodal big data model, thereby obtaining intermediate product information. Finally, the multimodal big data model can be used to process the intermediate product information to unify and standardize the product information from different sources, thereby obtaining the target product information for subsequent use.

[0061] This can be understood as follows: although original product information from different sources also possesses product attributes and category standards, the standards used differ due to the different data sources. Directly using these standards cannot guarantee the accuracy of subsequent processing. Therefore, the multimodal information understanding capabilities of a multimodal large-scale model can be leveraged to associate and map similar attributes, thereby achieving a unified goal. In other words, after obtaining the original product information, category mapping and attribute mapping are needed to map the original product information from different sources to a unified information structure, resulting in target product information with the same information structure for subsequent use. Attribute mapping can be understood as inputting the original product information into the model, then using prompts to instruct the model to predict which attribute information the product possesses from the attribute names under the input category and output the predictions, thus unifying and standardizing product information from different data sources.

[0062] For example, when Platform A needs to build product recommendation information for merchants, it can first obtain the original product information of different products from different data sources. This information can include the title, product image, sales volume, search volume, reviews, etc. of each product. Considering that different data sources correspond to different standards, each product may have different information structure descriptions. In order to unify and standardize the information, a multimodal model can be used to process the original product information of different products in a unified manner according to preset standard prompts. This will result in target product information divided according to the same data structure and standard. Subsequently, the target product information can be used to complete the construction of product recommendation information.

[0063] In summary, by using a multimodal large model to process the product information corresponding to the target category, information processing efficiency can be effectively improved, and the information structure can be unified, which facilitates subsequent analysis and use.

[0064] Step S204: Aggregate the target category products according to the target product information to obtain a target product set, and calculate the trend percentage of the target products and multiple product trend information contained in the target product set.

[0065] Specifically, after obtaining the target product information for each product, further steps are taken to accurately construct the product recruitment theme. This involves clustering products and calculating the proportion of each product within the set in each product trend information to analyze product trends, which can then be mapped to the recruitment theme direction. Therefore, target category products can be aggregated based on the target product information to obtain a target product set. Subsequently, the trend proportions of the target products within the target product set and multiple product trend information can be calculated. This allows analysis of trend proportions to determine which product trend information best aligns with recent sales trends. Following this, target product trend information that meets the desired trend proportions can be filtered out for use in generating recruitment theme information.

[0066] Specifically, the target product set refers to a collection of products with the same attributes obtained by grouping products in the target category according to the target product information. Correspondingly, product trend information refers to information composed of key attributes, selling point attributes, and core attribute knowledge pre-defined by the target platform regarding product sales trends. This information is used to analyze the trend of products on each product trend information to analyze potentially popular product themes in the future, and is used for subsequent generation of product recruitment theme information. Correspondingly, trend proportion refers to calculating the proportion of a product in the set relative to each product trend information among all products in the set. A higher proportion indicates that the product trend information more accurately represents the future sales trend of the product.

[0067] Furthermore, product aggregation can be achieved by comparing information. In this embodiment, the specific implementation is as follows:

[0068] The target product information is compared with the benchmark product information preset by the target platform, and multiple product sets are constructed based on the comparison results; the global transaction information corresponding to each of the multiple product sets is determined; the multiple product sets are sorted according to the global transaction information, and the target product set is selected from the sorted multiple product sets.

[0069] Specifically, benchmark product information refers to the standard product information set by the target platform. This can be understood as standard product information set according to specified product attributes, used to group products with similar attributes into a set. For example, benchmark product information can be set according to a specified IP address, so that products associated with that IP address form a set. Correspondingly, a product set specifically refers to a set of products corresponding to different benchmark product information. Correspondingly, global transaction information specifically refers to the transaction information obtained after statistically analyzing the transaction information of products within each product set.

[0070] Based on this, when constructing the target product set, the target product information corresponding to different products can be compared with the benchmark product information preset by the target platform to construct multiple product sets based on the comparison results. Then, considering that the product set with higher transaction volume can more accurately analyze product trends, the global transaction information corresponding to each of the multiple product sets can be determined. Then, the multiple product sets can be sorted according to the global transaction information, so that the target product set can be selected from the sorted product sets. For example, the product set with the highest global transaction information can be selected as the target product set for use in the subsequent construction of product theme information.

[0071] In practical applications, when constructing a target product set, considering that the target product information for each product may include product images, product titles, product attributes, etc., in order to ensure the accuracy of the set construction, image similarity calculation, product title similarity calculation, and product attribute similarity calculation can be used to select products from the target category products that are similar to the benchmark product information set by the platform to form a set, thereby ensuring that the products contained in each set are similar or identical products.

[0072] In summary, by using information comparison to aggregate products, it can be ensured that the products contained in the constructed set are similar or identical, thus facilitating subsequent trend information analysis.

[0073] Furthermore, determining the trend percentage corresponding to any one of the multiple product trend information sets includes:

[0074] Determine the target product attribute information corresponding to the target product in the target product set; calculate the number of products in the target product set that match the product trend information based on the target product attribute information and product trend information; calculate the trend percentage corresponding to the product trend information based on the number of products and the global number of products corresponding to the target product set.

[0075] Specifically, the target product attribute information refers to the attribute information corresponding to each target product in the target product set. Correspondingly, the product quantity refers to the total number of products in the target product set that match the product trend information, and the global product quantity refers to the total number of products contained in the target product set.

[0076] Based on this, when calculating the trend percentage corresponding to any product trend information, we can first extract the target product attribute information corresponding to each target product. Then, we can determine the matching degree of the trend information of each target product associated with the target product set by calculating the information matching degree. Then, by selecting products with a matching degree greater than a preset threshold as the products associated with the product trend information, we can count the number of products. Then, we can calculate the ratio of the number of products to the number of products in the target product set to obtain the trend percentage corresponding to the product trend information. By analogy, after obtaining the trend percentage corresponding to each product trend information, it can be used to complete the construction of product theme information in combination with the large language model.

[0077] In practical applications, considering that aggregation can yield multiple product sets, and that product trends can be represented by product sets with high transaction volumes or order numbers, target products can be selected from multiple product sets based on transaction information for use in constructing product recruitment theme information. Furthermore, based on the platform's preset key attributes, selling point attributes, and core attribute knowledge of leaf category dimensions, the CPV (Product Category or Product View) hit ratio in the target product set can be sorted, allowing for the selection of a predetermined amount of trend information for subsequent construction of product recruitment theme information.

[0078] Following the previous example, after preprocessing the product information for all products, the product information for each product can be compared with the platform's preset standard product information. Based on the comparison results, similar products can be selected to form multiple product sets. Then, the transaction volume of each product set can be statistically analyzed, and the product set with the highest transaction volume can be selected as the target product set. For example, if the target product set consists of women's clothing, the hit rate can be calculated based on the platform's preset core attributes for women's clothing: Style, Material, Fabric Type; key attributes: Dresses Length; and selling point attributes: Feature (Breathable, Dry Cleaning), etc. This yields the percentage of products within each set corresponding to each selling point attribute. Then, key selling point attributes can be selected from these percentages for subsequent use.

[0079] In summary, by using a calculated proportion to determine the trend proportion corresponding to each product's trend information, product trends can be subjectively mapped from a data perspective. Based on this, the construction of product theme information can be carried out, ensuring the accuracy of the construction of product theme information.

[0080] Step S206: Filter the target product trend information from the multiple product trend information according to the trend proportion.

[0081] Specifically, after obtaining the trend percentage corresponding to each product trend information as described above, the multiple product trend information can be sorted according to the trend percentage. The superiority of each product trend information can be determined by the sorting result. Therefore, a set number of product trend information can be selected as target product trend information based on the sorting result, so as to construct the product theme information in combination with the product trend information in the future.

[0082] In practice, multiple product sets can be obtained by aggregating products, and multiple product trend information can be obtained by calculating the proportion of trend information for each product set. For each product trend information, the corresponding product theme information can be constructed using the information processing method provided in this embodiment. This embodiment describes the process by taking the construction of the corresponding product theme information for the target product trend information as an example.

[0083] Based on this, the target product trend information obtained can be used as trend information such as trend categories, attribute items, and images for targeted product recruitment. For example, for women's clothing products, which include the Women's Day attribute and images with pink as the main color, product recruitment theme information can be generated according to this information. This information can then be used to recommend merchants to publish products based on this information, thereby increasing product sales.

[0084] Step S208: The target product trend information is processed using a large language model according to preset product suggestion words to obtain product theme information, wherein the product theme information is used to recommend product release strategies to target merchants.

[0085] Specifically, after obtaining the target product trend information, considering that the target product trend information is composed of multi-dimensional information and may not be convenient for merchants to browse, a large language model can be used to process the target product trend information according to the preset product prompts. This allows the downstream service to recommend product release strategies to target merchants based on the product theme information, thereby increasing the merchants' product sales.

[0086] Specifically, pre-configured product suggestion keywords refer to suggestion words pre-configured for the large language model. These suggestion words enable the model to organize target product trend information into thematic information that aligns with the merchant's reference for launching new products. Correspondingly, product suggestion thematic information refers to information provided to merchants for reference when launching new products, aiming to increase sales after the product launch. The product launch strategy refers to the strategies referenced when launching new products, such as launch time, product theme description, product introduction, and product endorsement analysis.

[0087] Furthermore, when processing target product trend information using a large language model, in order to generate product-specific information based on this information, pre-set product prompts can be used to help the model understand the operation that needs to be performed. In this embodiment, the specific implementation is as follows:

[0088] The target product trend information is input into the large language model, and the preset product prompt words of the large language model are determined; information to be processed is constructed based on the target product trend information and the preset product prompt words; the information to be processed is processed using the large language model to obtain product theme information.

[0089] Specifically, the information to be processed refers to the information obtained after processing the target product trend information using pre-set product suggestion words. Based on this, when processing the target product trend information using a large language model, the target product trend information can first be input into the large language model, while simultaneously determining the pre-set product suggestion words for the large language model; then, the information to be processed can be constructed based on the target product trend information and the pre-set product suggestion words; finally, the large language model can be used to process the information to be processed, thereby obtaining the product theme information, which can then be used to recommend products to merchants.

[0090] In summary, by combining pre-set product prompts to predict product theme information, the large language model can fully understand the current prediction needs, thereby improving the accuracy of theme information prediction.

[0091] After obtaining the product recommendation theme information, considering the large number of merchants on the platform, not all merchants are suitable for publishing products according to the product recommendation theme information. Therefore, target merchants can be filtered according to the product recommendation theme information. In this embodiment, the specific implementation method is as follows:

[0092] Obtain merchant profiles corresponding to multiple merchants associated with the target platform; match the product promotion theme information with the merchant profiles corresponding to the multiple merchants, and filter target merchants among the multiple merchants based on the matching results; send the product promotion theme information to the target merchants to recommend that the target merchants publish new products on the target platform according to the product promotion theme information.

[0093] Specifically, a merchant profile refers to a profile constructed by combining multi-dimensional information about a merchant, such as the type of goods sold, the price of the goods, and customer reviews. Correspondingly, a target merchant refers to a merchant among multiple merchants whose product information matches the advertised theme.

[0094] Based on this, in order to enable merchants matching the product theme information to quickly refer to the information, we can obtain the merchant profiles corresponding to multiple merchants associated with the target platform. Then, we can match the product theme information with the merchant profiles corresponding to multiple merchants, thereby filtering the target merchants among multiple merchants based on the matching results. After that, the product theme information can be sent to the target merchants to recommend that they publish new products on the target platform according to the product theme information.

[0095] In practical applications, since each merchant operates in different categories and market segments, after generating product recruitment theme information through a large language model, in order to avoid pushing a large number of invalid product recruitment theme information to merchants, the product recruitment theme information can be matched with the merchant profile. This allows for the push of potentially relevant product recruitment theme information to merchants based on their historical product launch information, browsing and clicking behavior, thereby improving product recruitment efficiency.

[0096] Following the previous example, after obtaining the target product trend information, the target product trend information can be input into the large language model for processing. This allows the large language model to generate product theme information according to the preset product prompts. The product theme information can include the product's theme and the reason for recommending the product, thus making the product theme information more interpretable. For example, if the product theme is "new women's coat", the reason for recommending the product is "With Women's Day approaching and the temperature dropping, women's coats are especially suitable for purchase recently, which can meet market demand and increase sales".

[0097] Furthermore, after obtaining the aforementioned product theme information, merchant profiles of merchants on the target platform can be acquired. Subsequently, by calculating the matching degree between the merchant profiles and the product theme information, merchants selling women's clothing can be selected from multiple merchants to recommend the aforementioned product theme information, thereby enabling such merchants to increase their product sales.

[0098] In summary, by matching merchant profiles with product theme information, merchants matching the product theme information can be selected from multiple merchants and the information can be sent to them. This helps merchants plan their product releases and improves their onboarding experience.

[0099] Furthermore, to enable the large language model to use more accurate prompt words for information processing, the pre-set prompt words can be determined through iterative optimization. In this embodiment, the specific implementation is as follows:

[0100] The large language model is used to process the sample product trend information according to the initial product suggestion words to obtain the predicted product theme information; the predicted product theme information and the sample product trend information are input into the evaluation model; the evaluation model evaluates the initial product suggestion words according to the sample product theme information corresponding to the sample product trend information to obtain the suggestion word optimization information corresponding to the initial product suggestion words; the initial product suggestion words are optimized according to the suggestion word optimization information until the preset product suggestion words that meet the model configuration conditions are obtained.

[0101] Specifically, the initial product suggestion words refer to the preset suggestion words for the large language model that are yet to be optimized. Correspondingly, the predicted product theme information refers to the product theme information obtained by the large language model after processing sample product trend information based on the initial product suggestion words. The evaluation model refers to the model that evaluates the output of the large language model in each round based on samples and labels, thereby determining the direction of suggestion word optimization. It evaluates the model's output based on samples and labels. Since the large language model is a trained model, inaccuracies in the model's output are influenced by the suggestion words; therefore, the evaluation model's result represents the optimization information for the suggestion words. Correspondingly, suggestion word optimization information refers to information on optimizing the suggestion words, such as adding, deleting, or modifying the content contained in the suggestion words. Correspondingly, model configuration conditions refer to the conditions for stopping the optimization of product suggestion words, including but not limited to the number of iterations and validation.

[0102] Based on this, in the prompt word optimization stage, a large language model can first be used to process the sample product trend information according to the initial product prompt words to obtain the predicted product theme information. Then, the predicted product theme information and sample product trend information can be input into the evaluation model. The evaluation model evaluates the initial product prompt words according to the sample product theme information corresponding to the sample product trend information, so as to obtain the prompt word optimization information corresponding to the initial product prompt words based on the evaluation results. On this basis, the initial product prompt words can be optimized according to the prompt word optimization information, and it can be checked whether the optimized product prompt words meet the model configuration conditions. If they do not meet the conditions, new samples can be selected to continue the above processing until a preset product prompt word that meets the model configuration conditions is obtained.

[0103] In summary, by constructing pre-set product prompts through iterative optimization, we can ensure that these prompts better meet current usage needs, thereby enabling the large language model to output more accurate topic information for downstream use.

[0104] Furthermore, when there are multiple product listing topics, the list can be updated according to a pre-defined strategy to facilitate browsing for merchants, allowing them to view real-time and accurate product listing topic information. In this embodiment, the specific implementation is as follows:

[0105] Obtain traffic information corresponding to multiple product theme information, sort and update the multiple product theme information according to the traffic information, generate and display a first product theme information list based on the sorting and update results; or, obtain time information corresponding to multiple product theme information, remove and update the multiple product theme information according to the time information, generate and display a second product theme information list based on the removal and update results.

[0106] Specifically, traffic information refers to the access traffic or exposure volume corresponding to each product topic information. Correspondingly, the first product topic information list is a list obtained by sorting multiple product topic information according to traffic information. Correspondingly, time information refers to the time when each product topic information was published. Correspondingly, the second product topic information list is a list obtained by sorting multiple product topic information according to time information.

[0107] Based on this, considering that the platform can generate multiple product listing topics based on the above processing, in order to provide merchants with an interface to view these topics, the multiple product listing topics can be sorted to obtain a list of merchant listing topics. Furthermore, to display the real-time effectiveness of the product listing topics, the information in the list can be sorted by traffic or by publication time.

[0108] In other words, it is possible to obtain traffic information corresponding to multiple product theme information, sort and update the multiple product theme information according to the traffic information, and then generate and display a first product theme information list based on the sorting and update results; or, it is possible to obtain time information corresponding to multiple product theme information, and then remove and update the multiple product theme information according to the time information, in order to remove product theme information with a long posting time, and then generate and display a second product theme information list based on the removal and update results.

[0109] In practice, refined traffic control can be implemented based on themes to give more exposure to themes urgently needing product recruitment. For example, themes can be stratified based on their source, and weighted according to priority to increase the exposure of high-priority themes; or they can be sorted by publication time to increase the exposure of newly published themes and decrease the exposure of older themes; or they can be sorted by the number of products recruited, reducing the exposure of themes once a certain number of products are recruited; or they can be sorted by click-through rate to increase the exposure of themes with high click-through rates. In practical applications, the above strategies can be set according to actual needs, and this embodiment does not impose too many limitations. Furthermore, for product recruitment themes that have been online for a period of time, metrics such as product recruitment scale, conversion rate, and order volume can be evaluated. If they do not meet expectations, they can be taken offline to prevent low-quality product recruitment themes from continuously guiding merchants to publish low-quality products.

[0110] The information processing method provided in this embodiment, in order to improve the efficiency of product information generation while ensuring accuracy, can first preprocess the original product information of the target category products in the target platform into target product information with the corresponding preset information structure of the target platform after obtaining the original product information of the target category products, thereby realizing the unification of the information structure of multi-source product information. Then, the target category products can be aggregated according to the target product information to obtain a target product set, so that products with the same attributes are grouped into the same set. On this basis, the trend proportions of the target products contained in the target product set and the trend information of multiple products can be calculated respectively, and the target product trend information can be filtered from the multiple product trend information according to the trend proportions; thus realizing the analysis of the proportions of existing products relative to the preset product trend information to sort out the product selling point trends. Afterwards, the large language model can be used to process the trend information of target products according to the preset product recommendation prompts to obtain product recommendation theme information. This enables the recommendation of product release strategies to target merchants based on the product recommendation theme information. By analyzing the target category products within the platform, product recommendation theme information can be automatically generated, which not only ensures the accuracy of information generation but also improves the efficiency of product recommendation on the platform. As a result, merchants can quickly complete targeted product recommendation based on the generated product recommendation theme information, thereby boosting the platform's transaction volume and improving the user's product purchasing experience.

[0111] Referring to Figure 3, Figure 3 shows a flowchart of another information processing method provided according to an embodiment of the present disclosure, which specifically includes the following steps.

[0112] Step S302: Obtain the original product information of the target category products in the target platform, and preprocess the original product information into target product information corresponding to the preset information structure of the target platform.

[0113] Step S304: Aggregate the target category products according to the target product information to obtain a target product set, and calculate the trend percentage of the target products and multiple product trend information contained in the target product set.

[0114] Step S306: Filter target product trend information from the multiple product trend information according to the trend proportion, and process the target product trend information using a large language model according to preset product prompt words to obtain product theme information.

[0115] Step S308: Select target merchants that match the product theme information from among the multiple merchants associated with the target platform, and recommend the product theme information to the target merchants.

[0116] The other information processing method provided in this embodiment corresponds to the information processing method described above. For any content not described in detail in this embodiment, please refer to the description in the above embodiments. This embodiment will not elaborate further here.

[0117] In summary, to improve the efficiency and accuracy of product information generation, after obtaining the original product information of the target category products on the target platform, the original product information can be preprocessed into target product information with the corresponding preset information structure of the target platform, thereby achieving a unified information structure for multi-source product information. Then, target category products can be aggregated based on the target product information to obtain a target product set, ensuring that products with the same attributes form the same set. Based on this, the trend proportions of the target products included in the target product set and multiple product trend information can be calculated, and the target product trend information can be filtered from multiple product trend information according to the trend proportions. This allows for the analysis of the proportions of existing products relative to preset product trend information to identify product selling point trends. Afterwards, the large language model can be used to process the trend information of target products according to the preset product recommendation prompts to obtain product recommendation theme information. This enables the recommendation of product release strategies to target merchants based on the product recommendation theme information. By analyzing the target category products within the platform, product recommendation theme information can be automatically generated, which not only ensures the accuracy of information generation but also improves the efficiency of product recommendation on the platform. As a result, merchants can quickly complete targeted product recommendation based on the generated product recommendation theme information, thereby boosting the platform's transaction volume and improving the user's product purchasing experience.

[0118] The following description, in conjunction with Figure 4, uses the application of the information processing method provided in this disclosure in a shopping platform as an example to further illustrate the information processing method. Figure 4 shows a flowchart of the processing procedure of an information processing method according to an embodiment of this disclosure, specifically including the following steps.

[0119] Step S402: Obtain product details, product attribute information, and product transaction information of the target category products in the target platform, and generate original product information based on the product details, product attribute information, and product transaction information.

[0120] Step S404: Input the original product information into the multimodal big model, and update the original product information with the standard prompt words corresponding to the preset information structure of the target platform in the multimodal big model to obtain intermediate product information.

[0121] Step S406: Process the intermediate product information using a multimodal large model to obtain the target product information.

[0122] Step S408: Compare the target product information with the benchmark product information preset by the target platform, and construct multiple product sets based on the comparison results.

[0123] Step S410: Determine the global transaction information corresponding to the multiple product sets, sort the multiple product sets according to the global transaction information, and select the target product set from the sorted multiple product sets.

[0124] Step S412: Calculate the trend percentage of the target product and multiple product trend information contained in the target product set.

[0125] The determination of the trend percentage corresponding to any one of the multiple product trend information includes: determining the target product attribute information corresponding to the target product in the target product set; calculating the number of products matching the product trend information in the target product set based on the target product attribute information and the product trend information; and calculating the trend percentage corresponding to the product trend information based on the number of products and the global number of products corresponding to the target product set.

[0126] Step S414: Sort multiple product trend information according to trend proportion, and select a set number of product trend information as target product trend information based on the sorting result.

[0127] Step S416: Input the target product trend information into the big language model and determine the preset product prompt words of the big language model.

[0128] Step S418: Construct information to be processed based on the target product trend information and preset product prompts, and use a large language model to process the information to be processed to obtain product theme information.

[0129] Step S420: Obtain the merchant profiles corresponding to multiple merchants associated with the target platform, match the product theme information with the merchant profiles corresponding to multiple merchants, and filter the target merchants from among the multiple merchants based on the matching results.

[0130] Step S422: Send the product theme information to the target merchant to recommend that the target merchant publish new products on the target platform according to the product theme information.

[0131] In summary, to improve the efficiency and accuracy of product information generation, after obtaining the original product information of the target category products on the target platform, the original product information can be preprocessed into target product information with the corresponding preset information structure of the target platform, thereby achieving a unified information structure for multi-source product information. Then, target category products can be aggregated based on the target product information to obtain a target product set, ensuring that products with the same attributes form the same set. Based on this, the trend proportions of the target products included in the target product set and multiple product trend information can be calculated, and the target product trend information can be filtered from multiple product trend information according to the trend proportions. This allows for the analysis of the proportions of existing products relative to preset product trend information to identify product selling point trends. Afterwards, the large language model can be used to process the trend information of target products according to the preset product recommendation prompts to obtain product recommendation theme information. This enables the recommendation of product release strategies to target merchants based on the product recommendation theme information. By analyzing the target category products within the platform, product recommendation theme information can be automatically generated, which not only ensures the accuracy of information generation but also improves the efficiency of product recommendation on the platform. As a result, merchants can quickly complete targeted product recommendation based on the generated product recommendation theme information, thereby boosting the platform's transaction volume and improving the user's product purchasing experience.

[0132] Corresponding to the above method embodiments, this disclosure also provides an information processing device embodiment. Figure 5 shows a schematic diagram of the structure of an information processing device provided in one embodiment of this disclosure. As shown in Figure 5, the device includes:

[0133] The acquisition module 502 is configured to acquire the original product information of the target category products in the target platform, and preprocess the original product information into target product information corresponding to the preset information structure of the target platform;

[0134] The aggregation module 504 is configured to aggregate the target category products according to the target product information to obtain a target product set, and calculate the trend proportions of the target products and multiple product trend information contained in the target product set respectively.

[0135] The filtering module 506 is configured to filter target product trend information from the plurality of product trend information according to the trend proportion;

[0136] The processing module 508 is configured to use a large language model to process the target product trend information according to preset product prompts to obtain product theme information, wherein the product theme information is used to recommend product release strategies to target merchants.

[0137] In an optional embodiment, obtaining the original product information of the target category products in the target platform and preprocessing the original product information into target product information corresponding to the preset information structure of the target platform includes:

[0138] The system acquires product details, product attributes, and product transaction information for a target category of goods on a target platform, and generates original product information based on these information. It then inputs the original product information into a multimodal big data model, updates it using standardized prompts corresponding to the target platform's preset information structure within the model, and obtains intermediate product information. Finally, it processes the intermediate product information using the multimodal big data model to obtain the target product information.

[0139] In an optional embodiment, the step of aggregating the target category products based on the target product information to obtain a target product set includes:

[0140] The target product information is compared with the benchmark product information preset by the target platform, and multiple product sets are constructed based on the comparison results; the global transaction information corresponding to each of the multiple product sets is determined; the multiple product sets are sorted according to the global transaction information, and the target product set is selected from the sorted multiple product sets.

[0141] In an optional embodiment, determining the trend percentage corresponding to any one of the plurality of product trend information includes:

[0142] The process involves: determining the target product attribute information corresponding to the target product in the target product set; calculating the number of products in the target product set that match the product trend information based on the target product attribute information and product trend information; calculating the trend percentage corresponding to the product trend information based on the number of products and the global number of products corresponding to the target product set; wherein, the step of filtering target product trend information from multiple product trend information according to the trend percentage includes: sorting the multiple product trend information according to the trend percentage, and selecting a set number of product trend information as target product trend information based on the sorting result.

[0143] In an optional embodiment, the step of processing the target product trend information using a large language model according to preset product prompts to obtain product theme information includes:

[0144] The target product trend information is input into the large language model, and the preset product prompt words of the large language model are determined; information to be processed is constructed based on the target product trend information and the preset product prompt words; the information to be processed is processed using the large language model to obtain product theme information.

[0145] In an optional embodiment, after the step of processing the target product trend information using a large language model according to preset product prompts to obtain product theme information, the method further includes:

[0146] Obtain merchant profiles corresponding to multiple merchants associated with the target platform; match the product promotion theme information with the merchant profiles corresponding to the multiple merchants, and filter target merchants among the multiple merchants based on the matching results; send the product promotion theme information to the target merchants to recommend that the target merchants publish new products on the target platform according to the product promotion theme information.

[0147] In an optional embodiment, when there are multiple product theme information items, after the step of processing the target product trend information using a large language model according to preset product prompt words to obtain product theme information is executed, the method further includes:

[0148] Obtain traffic information corresponding to multiple product theme information, sort and update the multiple product theme information according to the traffic information, generate and display a first product theme information list based on the sorting and update results; or, obtain time information corresponding to multiple product theme information, remove and update the multiple product theme information according to the time information, generate and display a second product theme information list based on the removal and update results.

[0149] In an optional embodiment, determining the preset product prompts includes:

[0150] The large language model is used to process the sample product trend information according to the initial product suggestion words to obtain the predicted product theme information; the predicted product theme information and the sample product trend information are input into the evaluation model; the evaluation model evaluates the initial product suggestion words according to the sample product theme information corresponding to the sample product trend information to obtain the suggestion word optimization information corresponding to the initial product suggestion words; the initial product suggestion words are optimized according to the suggestion word optimization information until the preset product suggestion words that meet the model configuration conditions are obtained.

[0151] The information processing device provided in this embodiment, in order to improve the efficiency of product information generation while ensuring accuracy, can, after obtaining the original product information of target category products in the target platform, first preprocess the original product information into target product information corresponding to the preset information structure of the target platform, thereby achieving a unified information structure for multi-source product information. Then, it can aggregate target category products based on the target product information to obtain a target product set, so that products with the same attributes form the same set. Based on this, it can calculate the trend proportion of the target products included in the target product set and the trend information of multiple products, and filter the target product trend information from the multiple product trend information according to the trend proportion; thus realizing the analysis of the proportion of existing products relative to the preset product trend information to organize the product selling point trend. Afterwards, the large language model can be used to process the trend information of target products according to the preset product recommendation prompts to obtain product recommendation theme information. This enables the recommendation of product release strategies to target merchants based on the product recommendation theme information. By analyzing the target category products within the platform, product recommendation theme information can be automatically generated, which not only ensures the accuracy of information generation but also improves the efficiency of product recommendation on the platform. As a result, merchants can quickly complete targeted product recommendation based on the generated product recommendation theme information, thereby boosting the platform's transaction volume and improving the user's product purchasing experience.

[0152] The above is an illustrative scheme of an information processing device according to this embodiment. It should be noted that the technical solution of this information processing device and the technical solution of the information processing method described above belong to the same concept. For details not described in detail in the technical solution of the information processing device, please refer to the description of the technical solution of the information processing method described above.

[0153] Corresponding to the above method embodiments, this disclosure also provides another information processing apparatus embodiment. FIG6 shows a schematic diagram of the structure of another information processing apparatus provided in an embodiment of this disclosure. As shown in FIG6, the apparatus includes:

[0154] The information acquisition module 602 is configured to acquire the original product information of the target category products in the target platform, and preprocess the original product information into target product information corresponding to the preset information structure of the target platform;

[0155] The product aggregation module 604 is configured to aggregate the target category products based on the target product information to obtain a target product set, and calculate the trend percentage of the target products and multiple product trend information contained in the target product set respectively.

[0156] The filtering information module 606 is configured to filter target product trend information from multiple product trend information according to the trend proportion, and process the target product trend information using a large language model according to preset product prompt words to obtain product theme information;

[0157] The merchant matching module 608 is configured to filter target merchants that match the product theme information from multiple merchants associated with the target platform, and recommend the product theme information to the target merchants.

[0158] The above is an illustrative scheme of another information processing device according to this embodiment. It should be noted that the technical solution of this other information processing device and the technical solution of the other information processing method described above belong to the same concept. For details not described in detail in the technical solution of the other information processing device, please refer to the description of the technical solution of the other information processing method described above.

[0159] Figure 7 shows a structural block diagram of a computing device 700 according to an embodiment of the present disclosure. The components of the computing device 700 include, but are not limited to, a memory 710 and a processor 720. The processor 720 is connected to the memory 710 via a bus 730, and a database 750 is used to store data.

[0160] The computing device 700 also includes an access device 740, which enables the computing device 700 to communicate via one or more networks 760. Examples of these networks include Public Switched Telephone Network (PSTN), Local Area Network (LAN), Wide Area Network (WAN), Personal Area Network (PAN), or combinations of communication networks such as the Internet. The access device 740 may include one or more of any type of wired or wireless network interface (e.g., a network interface controller (NIC)), such as an IEEE 802.11 Wireless Local Area Network (WLAN) wireless interface, a Wi-MAX (Worldwide Interoperability for Microwave Access) interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, or a Near Field Communication (NFC) interface.

[0161] In one embodiment of this disclosure, the aforementioned components of the computing device 700, as well as other components not shown in FIG. 7, may be interconnected, for example, via a bus. It should be understood that the computing device block diagram shown in FIG. 7 is merely for illustrative purposes and is not intended to limit the scope of this disclosure. Those skilled in the art can add or replace other components as needed.

[0162] The computing device 700 can be any type of stationary or mobile computing device, including mobile computers or mobile computing devices (e.g., tablet computers, personal digital assistants, laptop computers, notebook computers, netbooks, etc.), mobile phones (e.g., smartphones), wearable computing devices (e.g., smartwatches, smart glasses, etc.) or other types of mobile devices, or stationary computing devices such as desktop computers or personal computers (PCs). The computing device 700 can also be a mobile or stationary server.

[0163] The processor 720 is configured to execute the following computer-executable instructions, which, when executed by the processor, implement the steps of the above-described information processing method.

[0164] The above is an illustrative scheme of a computing device according to this embodiment. It should be noted that the technical solution of this computing device and the technical solution of the information processing method described above belong to the same concept. For details not described in detail in the technical solution of the computing device, please refer to the description of the technical solution of the information processing method described above.

[0165] An embodiment of this disclosure also provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the above-described information processing method.

[0166] The above is an illustrative scheme of a computer-readable storage medium according to this embodiment. It should be noted that the technical solution of this storage medium and the technical solution of the information processing method described above belong to the same concept. For details not described in detail in the technical solution of the storage medium, please refer to the description of the technical solution of the information processing method described above.

[0167] An embodiment of this disclosure also provides a computer program, wherein when the computer program is executed in a computer, it causes the computer to perform the steps of the above-described information processing method.

[0168] The above is an illustrative example of a computer program according to this embodiment. It should be noted that the technical solution of this computer program and the technical solution of the information processing method described above belong to the same concept. Details not described in detail in the technical solution of the computer program can be found in the description of the technical solution of the information processing method described above.

[0169] An embodiment of this disclosure also provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described information processing method.

[0170] The above is an illustrative scheme of a computer program product according to this embodiment. It should be noted that the technical solution of this computer program product and the technical solution of the information processing method described above belong to the same concept. For details not described in detail in the technical solution of the computer program product, please refer to the description of the technical solution of the information processing method described above.

[0171] The foregoing has described specific embodiments of this disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired results. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0172] The computer instructions include computer program code, which may be in the form of source code, object code, executable file, or certain intermediate forms. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium may be appropriately added or removed according to the requirements of patent practice. For example, in some regions, according to patent practice, computer-readable media may not include electrical carrier signals and telecommunication signals.

[0173] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of this disclosure are not limited to the described order of actions, because according to the embodiments of this disclosure, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the embodiments of this disclosure.

[0174] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0175] The preferred embodiments disclosed above are merely illustrative of this disclosure. The optional embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the embodiments of this disclosure. These embodiments are selected and specifically described in this disclosure to better explain the principles and practical applications of the embodiments of this disclosure, thereby enabling those skilled in the art to better understand and utilize this disclosure. This disclosure is limited only by the claims and their full scope and equivalents.

Claims

1. An information processing method, comprising: obtaining original product information of target category products in a target platform, and preprocessing the original product information into target product information corresponding to a preset information structure of the target platform; aggregating the target category products according to the target product information, obtaining a target product set, and calculating a trend proportion corresponding to each of a plurality of product trend information respectively for a target product included in the target product set; screening target product trend information from the plurality of product trend information according to the trend proportion; processing the target product trend information according to a preset product recruitment prompt word by using a large language model to obtain product recruitment theme information, wherein the product recruitment theme information is used to recommend a product publishing strategy to a target merchant.

2. The information processing method of claim 1, wherein the obtaining original product information of target category products in a target platform, and preprocessing the original product information into target product information corresponding to a preset information structure of the target platform comprises: obtaining product detail information, product attribute information, and product transaction information of target category products in a target platform, and generating original product information according to the product detail information, the product attribute information, and the product transaction information; inputting the original product information into a multi-modal large model, updating the original product information by using a standard prompt word corresponding to the preset information structure of the target platform in the multi-modal large model to obtain intermediate product information; and processing the intermediate product information by using the multi-modal large model to obtain target product information.

3. The information processing method of claim 1 or 2, wherein the aggregating the target category products according to the target product information to obtain a target product set comprises: comparing the target product information with reference product information preset by the target platform, and constructing a plurality of product sets according to a comparison result; determining global transaction information corresponding to each of the plurality of product sets; and sorting the plurality of product sets according to the global transaction information, and selecting a target product set from the sorted plurality of product sets. The comparison of the target product information with the reference product information preset by the target platform comprises:

4. The information processing method according to claim 3, wherein comprehensively determining whether the products belong to the same product set based on visual similarity of product pictures, semantic similarity of product titles, and structured matching degree of product attributes.

5. The information processing method of any one of claims 1 to 4, wherein the determination of a trend proportion corresponding to any one of the plurality of product trend information comprises: determining target product attribute information corresponding to a target product in the target product set; calculating a number of products in the target product set that match the product trend information according to the target product attribute information and the product trend information; calculating a trend proportion corresponding to the product trend information based on the number of products and a global number of products corresponding to the target product set; and wherein the screening of target product trend information from the plurality of product trend information according to the trend proportion comprises: ​ According to the trend proportion, the plurality of commodity trend information is sorted, and a set number of commodity trend information is selected as target commodity trend information according to the sorting result.

6. The information processing method of any one of claims 1 to 5, wherein the processing of the target commodity trend information according to the preset product promotion prompt word by the large language model to obtain product promotion theme information comprises: inputting the target commodity trend information into the large language model and determining a preset product promotion prompt word of the large language model; constructing to-be-processed information according to the target commodity trend information and the preset product promotion prompt word; processing the to-be-processed information by the large language model to obtain product promotion theme information.

7. The information processing method according to claim 6, wherein The preset product promotion prompt word contains product promotion scene context, target user portrait features, and seasonal marketing elements, and is used to guide the large language model to generate product promotion theme information with timeliness and market adaptability.

8. The information processing method of any one of claims 1 to 7, wherein after the step of processing the target commodity trend information according to the preset product promotion prompt word by the large language model to generate product promotion theme information, the method further comprises: obtaining a plurality of merchant portraits corresponding to a plurality of merchants associated with the target platform; matching the product promotion theme information with the merchant portraits corresponding to the plurality of merchants, and filtering target merchants from the plurality of merchants according to a matching result; sending the product promotion theme information to the target merchants, so as to recommend the target merchants to publish new commodities on the target platform according to the product promotion theme information.

9. The information processing method according to claim 8, wherein The merchant portrait includes category distribution, price band interval, average conversion rate, and user evaluation score of historical published commodities of the merchant; and the matching process is implemented by vector similarity or rule engine to align the product promotion theme information with the merchant's business capability.

10. The information processing method of claim 1, wherein after the step of processing the target commodity trend information according to the preset product promotion prompt by the large language model to obtain product promotion theme information, when there are a plurality of product promotion theme information, the method further comprises: obtaining traffic information corresponding to a plurality of product promotion theme information, updating the plurality of product promotion theme information according to the traffic information, generating a first product promotion theme information list according to an updating result, and displaying the first product promotion theme information list; or obtaining time information corresponding to a plurality of product promotion theme information, updating the plurality of product promotion theme information by excluding according to the time information, generating a second product promotion theme information list according to an updating result, and displaying the second product promotion theme information list. The updating by excluding according to the time information comprises:

11. The information processing method according to claim 10, wherein if the product promotion theme information is published for more than a preset valid period threshold and the corresponding product promotion commodity quantity does not reach a preset scale, the product promotion theme information is automatically removed from the second product promotion theme information list.

12. The information processing method of any one of claims 1 to 11, wherein the determination of the preset product promotion prompt word comprises: processing sample commodity trend information according to an initial product promotion prompt word by the large language model to obtain predicted product promotion theme information; inputting the predicted product promotion theme information and the sample commodity trend information into an evaluation model; and determining the preset product promotion prompt word according to an evaluation result of the evaluation model. The initial product prompt word is evaluated according to sample product theme information corresponding to the sample product trend information through the evaluation model, and prompt word optimization information corresponding to the initial product prompt word is obtained; The initial product prompt word is optimized according to the prompt word optimization information until the preset product prompt word satisfying the model configuration condition is obtained.

13. The information processing method according to any one of claims 1 to 12, wherein The product trend information includes key attributes, selling point attributes and core attribute knowledge under the leaf category dimension preset by the target platform; and the calculation of the trend proportion is based on the ratio of the number of products in the target product set that hit the key attributes, selling point attributes or core attributes to the total number of products.

14. The information processing method according to any one of claims 1 to 13, wherein After obtaining the product theme information, the corresponding product task card is generated based on the product theme information, and is pushed to the operation background of the target merchant; the product task card contains a theme name, a recommendation reason, an example product link and a one-key publishing entrance.

15. An information processing method, comprising: obtaining original product information of target category products in a target platform, and preprocessing the original product information into target product information corresponding to a preset information structure of the target platform; aggregating the target category products according to the target product information to obtain a target product set, and calculating trend proportions of target products included in the target product set corresponding to a plurality of product trend information respectively; screening target product trend information from the plurality of product trend information according to the trend proportions, and processing the target product trend information according to a preset product prompt word by using a large language model to obtain product theme information; screening target merchants matching the product theme information from a plurality of merchants associated with the target platform, and recommending the product theme information to the target merchants.

16. An information processing apparatus, comprising: an acquisition module configured to obtain original product information of target category products in a target platform, and preprocess the original product information into target product information corresponding to a preset information structure of the target platform; an aggregation module configured to aggregate the target category products according to the target product information to obtain a target product set, and calculate trend proportions of target products included in the target product set corresponding to a plurality of product trend information; a screening module configured to screen target product trend information from the plurality of product trend information according to the trend proportions; a processing module configured to process the target product trend information according to a preset product prompt word by using a large language model to generate product theme information, wherein the product theme information is used to recommend product publishing strategies to target merchants.

17. An information processing apparatus, comprising: an acquisition information module configured to obtain original product information of target category products in a target platform, and preprocess the target product information into target product information corresponding to a preset information structure of the target platform; The aggregation commodity module is configured to aggregate the target category commodities according to the target commodity information, obtain a target commodity set, and calculate a trend proportion of a target commodity included in the target commodity set and corresponding to each of the plurality of commodity trend information; The screening information module is configured to screen target commodity trend information from the plurality of commodity trend information according to the trend proportion, and process the target commodity trend information according to a preset product recruitment prompt word by using a large language model to obtain product recruitment theme information; The matching merchant module is configured to screen a target merchant matching the product recruitment theme information from a plurality of merchants associated with the target platform, and recommend the product recruitment theme information to the target merchant.

18. A computing device comprising: a memory and a processor; the memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions, and the computer executable instructions, when executed by the processor, implement the steps of the method in any one of claims 1 to 15.

19. A computer readable storage medium storing computer executable instructions, and the computer executable instructions, when executed by a processor, implement the steps of the method in any one of claims 1 to 15.

20. A computer program product comprising a computer program or instructions, and the computer program or instructions, when executed by a processor, implement the steps of the method in any one of claims 1 to 15.

Images