Method and Electronic Device for Providing Commodity Object Information

Abstract

Embodiments of the present application disclose a method and an electronic device for providing commodity object information. The method includes: determining at least one target commodity object to be provided to a target user and its original description information; determining the national / regional attribute information of the target user; performing conversion processing related to localization expression preferences on the original description information according to the national / regional attribute information of the target user to generate target description information; providing the target description information corresponding to the at least one target commodity object to the client corresponding to the target user, so as to provide the target description information through a target page, so that when providing information of the same commodity object to users with different national / regional attribute information through the target page, different target description information generated based on different localization expression preferences is provided. Through the embodiments of the present application, it is possible to achieve "one thousand people, one thousand faces" in the expression of the front-end description information of the commodity objects provided on the target page, improve the user experience, and is conducive to improving indicators such as click-through rate and conversion rate.

Description

Technical Field

[0001] This application relates to the field of information technology, and particularly to a method for providing product object information and an electronic device. Background Art

[0002] In a product information service system, "one product for one person" usually refers to formulating different product recommendation strategies for different consumers to achieve better sales results. Traditional product recommendations are based on the attributes of the products themselves. For example, products are recommended according to factors such as product category, price, and sales volume. However, "one product for one person" is consumer-based, that is, products are recommended according to factors such as consumers' interests, purchase history, and behaviors.

[0003] This "one product for one person" approach has a positive effect on improving product click-through rates, conversion rates, etc. However, how to further improve the user experience and indicators such as click-through rates and conversion rates has always been the focus of attention of those skilled in the art. Summary of the Invention

[0004] This application provides a method for providing product object information and an electronic device, which can enable the product objects provided on the target page to achieve "one product for one person" in the expression of front-end description information, improve the user experience, and is conducive to improving indicators such as click-through rates and conversion rates.

[0005] This application provides the following solutions:

[0006] A method for providing product object information, comprising:

[0007] Determine at least one target product object to be provided to a target user through a target page and its original description information;

[0008] Determine the national / regional attribute information of the target user;

[0009] According to the national / regional attribute information of the target user, perform conversion processing on the original description information related to the local expression preference to generate target description information; wherein, the original description information includes: the original title text content of the target product object, and the conversion processing related to the local expression preference includes: according to the attribute preference information of the products in the category to which the target product object belongs for the population corresponding to the national / regional attribute information, perform conversion processing on the text content related to the product attribute expression in the original title text;

[0010] Provide the target description information corresponding to the at least one target commodity object to the client corresponding to the target user, so as to provide the target description information through the target page, so that when providing information about the same commodity object to users with different national / regional attribute information through the target page, different target description information generated based on different localization expression preferences is provided.

[0011] Among them, the target commodity object is associated with multiple stock - keeping units (SKUs), and different SKUs correspond to different specification parameter values;

[0012] The text content related to the expression of commodity attributes in the original title text includes: the content of the specification parameter values corresponding to one of the SKUs included in the original title text;

[0013] The conversion process for the text content related to the expression of commodity attributes in the original title text includes:

[0014] According to the preference information of the population corresponding to the national / regional attribute information for the specification parameter values of the commodities in the category to which the target commodity object belongs, convert the content of the specification parameter values in the original title text into the content of the corresponding specification parameter values of other SKUs.

[0015] Among them, the conversion process for the original description information related to the localization expression preference further includes:

[0016] Convert the keywords related to the commodity name and / or adjectives included in the content of the original title text into the commonly used local words corresponding to the national / regional attribute information.

[0017] Among them, the related process of localizing the expression of the original description information to generate the target description information further includes:

[0018] Use an artificial intelligence (AI) large - scale parameter model to perform processing that conforms to the local grammar expression habits and / or context coherence processing on the title text content after the localization expression conversion to generate the target description information.

[0019] A method for providing commodity object information includes:

[0020] Determine the target commodity object to be provided to the target user through the target page and its original description information, where the target page includes the commodity details page of the target commodity object;

[0021] Determine the national / regional attribute information of the target user;

[0022] Based on the country / region-specific attribute information of the target user, perform conversion processing related to localization expression preferences on the original description information to generate target description information; wherein, the conversion processing related to localization expression preferences includes: if the target commodity object is associated with multiple SKUs corresponding to different attribute values / specification parameter values, when presenting the SKU selection interface through the commodity details page, reorder the multiple SKUs so as to preferentially display the SKUs corresponding to the attribute values / parameter values commonly used in localization corresponding to the country / region-specific attribute information.

[0023] A method for providing commodity object information, including:

[0024] Determine at least one target commodity object to be provided to the target user through the target page and its original description information;

[0025] Determine the country / region-specific attribute information of the target user;

[0026] Based on the country / region-specific attribute information of the target user, perform conversion processing related to localization expression preferences on the original description information to generate target description information; wherein, the original description information includes: the original rich media information of the target commodity object; the conversion processing related to localization expression preferences includes: using an AI large-scale parameter model, according to the country / region-specific attribute information of the target user, perform conversion processing on the original rich media information of the target commodity object for localization expression to generate target rich media information, so as to display the target rich media information on the target page.

[0027] Wherein, the original rich media information includes original image information;

[0028] The conversion processing related to localization expression preferences on the original description information includes:

[0029] According to the country / region-specific attribute information of the target user, perform conversion processing on the composition style, model character type, and / or atmosphere elements of the original image information to generate target image information that conforms to the localization preferences corresponding to the country / region-specific attribute information.

[0030] Wherein, the original rich media information includes original audio information;

[0031] The conversion processing related to localization expression preferences on the original description information includes:

[0032] Convert the original audio information according to the country / region attribute information of the target user to generate target audio information that conforms to the localization preferences corresponding to the country / region attribute information.

[0033] A computer-readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the steps of the method described in any one of the foregoing are implemented.

[0034] An electronic device, comprising:

[0035] One or more processors; and

[0036] A memory associated with the one or more processors, the memory being used to store program instructions, and when the program instructions are read and executed by the one or more processors, the steps of the method described in any one of the foregoing are executed.

[0037] According to the specific embodiments provided by the present application, the present application discloses the following technical effects:

[0038] Through the embodiments of the present application, when it is necessary to provide information about a target commodity object to a target user, the original description information of such a target commodity object and the country / region attribute information of the target user can be obtained first. Then, according to the country / region attribute information of the target user, relevant processing for localizing the original description information can be performed to generate target description information, so as to provide such target description information to the target user through a target page. In this way, it can be made that the expression of the commodity object provided on the target page realizes "one thousand people, one thousand faces" in the front-end description information, and is more in line with the localization preferences of the country / region to which the current target user belongs, thus making it easier to stimulate the user's shopping enthusiasm, improve the user experience, and further contribute to improving indicators such as click-through rate and conversion rate.

[0039] Of course, it is not necessary for any product implementing the present application to achieve all the above-mentioned advantages simultaneously. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application, and for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0041] Figure 1 is a schematic diagram of the system architecture provided by the embodiments of the present application;

[0042] Figure 2 is a flowchart of the server-side method provided by the embodiments of the present application;

[0043] Figure 3 is a flowchart of the client method provided by an embodiment of the present application;

[0044] Figure 4 is a schematic diagram of the server device provided by an embodiment of the present application;

[0045] Figure 5 is a schematic diagram of the client device provided by an embodiment of the present application;

[0046] Figure 6 is a schematic diagram of the electronic device provided by an embodiment of the present application. Detailed implementation manners

[0047] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art belong to the scope protected by the present application.

[0048] First of all, the inventors of the present application found in the process of implementing the embodiments of the present application that currently, the "one product for one person" recommendation in the e-commerce field has been relatively mature. However, it mainly realizes "one product for one person" from the perspective of product matching. For different users, the front-end expression of the same product is the same (including information such as title, description diagram, product introduction, etc. Although there are also methods for generating product pictures, they mainly highlight the benefits and have a single usage scenario). However, such a "consistent" expression method cannot stimulate the shopping enthusiasm of different users. Especially for the cross-border e-commerce scenario, the cultures, habits, and preferences of users in different countries / regions vary greatly. At this time, if personalized processing of country / region localization is also carried out on the front-end expression of product information, it is beneficial to further improve the user experience, and then improve indicators such as click-through rate and conversion rate. For example, in terms of habitual terms, it is usually called "video" in region A, while it is usually called "video news" in region B; "oranges" in region A are called "liucheng" in region C; "power banks" in region A are called "urine bags" in region B, and so on. In addition, in terms of visual presentation, users in different countries / regions also have different preferences. For example, Brazilian users generally like more enthusiastic and colorful compositions, while users in the European and American regions generally like simple expressions; in terms of product attribute / parameter information, there are also different expressions in different countries / regions. For example, for shoe sizes, the US uses size 6, corresponding to Chinese size 235 and European size 37; in terms of product model figures, users in the Asian region are more accustomed to seeing Asian models with a slender body type, while Europeans and Americans are more accustomed to seeing European and American models with a plump body type, and so on. These expression differences actually have a great impact on the user experience, and using matching expressions can gain more recognition from local people.

[0049] Therefore, in the embodiments of the present application, corresponding solutions are provided. In this solution, in scenarios such as determining the products to be recommended to users or returning products that meet the search criteria according to the user's search request, the front-end expression of the product description information can also be converted so that the converted description information is more in line with the localized expression habits of the current target user's national / regional attributes, and then provided to the current target user through the target page, so that the target user can have a more cordial feeling.

[0050] Among them, specifically when converting the description information of the front-end expression, it can include the conversion of description information in multiple modalities such as text information, attribute parameter information, rich media information (including images, audio, etc.), and convert it into description information that is more suitable for the localized expression of the user's location. Specifically, in a preferred implementation manner, the capabilities of relevant models such as an AI (Artificial Intelligence) model can also be used to achieve the localized conversion processing of multi-modal product description information, and make the converted description information smoother and more natural, avoiding problems such as unsmooth sentences caused by simple keyword replacement and other methods, and the processing efficiency is also higher, so that the conversion process can also be completed within the time range of hundreds of milliseconds, thus better supporting the implementation of the entire solution.

[0051] Since in the solution provided by the embodiments of the present application, the conversion process of description information in multiple modalities can be completed through an AI model, for better understanding, the relevant concepts of the AI model (especially the large-scale AI parameter model, which will be mainly introduced by taking the large-scale AI parameter model as an example hereinafter) will be briefly introduced below. The large-scale AI parameter model can also be called the large AI model, and it can refer to a type of foundation model. Specifically, it can refer to a model with a huge number of parameters trained using a large amount of data and capable of adapting to a series of downstream tasks. For the large AI model, not only does it have the characteristic of a huge number of parameters in terms of parameter scale (as the model is continuously iterated, the number of parameters usually also increases exponentially, from hundreds of millions to trillions, and then to quadrillions, or even more), but also from the perspective of modality support, the large AI model has gradually developed from supporting single tasks in single modalities such as pictures, images, text, voice, and video to supporting multiple tasks in multiple modalities. That is to say, large models usually also have the efficient understanding ability of multiple modality information, cross-modal perception ability, and the migration and execution ability of cross-differentiated tasks, etc., and may even have the multi-modal information perception ability similar to that of the human brain.

[0052] From another perspective, the AI large model is short for "Artificial Intelligence Pre-trained Large Model", which contains two meanings: "pre-training" and "large model". The combination of the two generates a new artificial intelligence mode, that is, after the model is pre-trained on a large-scale dataset, it can support various downstream applications without fine-tuning or only with a small amount of data fine-tuning. That is to say, thanks to its "large-scale pre-training + fine-tuning" paradigm, the AI large model can well adapt to different downstream tasks and demonstrate its strong versatility. Such a versatile AI large model, under the condition of sharing parameters, can obtain excellent performance only by making corresponding fine-tuning in different downstream application scenarios, breaking through the limitation that traditional AI models are difficult to generalize to other tasks.

[0053] From the perspective of the processing result, the above-mentioned AI large model also belongs to a generative model (Generative Model). Such models can not only predict results based on features, but also "understand" how the data is generated and "create" new data based on this.

[0054] With the support of the above capabilities and existing knowledge of the AI large model, the localization conversion of commodity description information in the embodiments of the present application can be better implemented. For example, specifically, the original description information of the commodity and the national / regional attribute information of the current target user can be input into the AI large model, and the AI large model can perform corresponding keyword conversion, attribute / parameter information conversion, image conversion, etc. In addition, for text content, it is also possible to process the localized grammar expression habits based on the converted keywords, etc., or perform context coherence processing, so that the converted text conforms more to the localized grammar expression habits of the current target user and is also more smooth, avoiding problems such as unsmooth sentences after direct keyword replacement, etc.

[0055] From the perspective of the system architecture, see Figure 1 , the embodiments of the present application can be implemented in a commodity information service system, which can specifically include a client and a server. Among them, the client is mainly used for displaying the front-end page, interacting with users, etc., and the server is mainly used to provide specific data. In the embodiments of the present application, the specific AI model, etc. can be saved on the server, and after the specific description information conversion processing is completed on the server, it is returned to the client for display and other processing. Of course, in actual applications, the computing resources of "end + cloud" can be more fully utilized. For some simple logics, the computing resources on the client side can be used for end-side computing, and complex logics can be computed on the cloud server, etc.

[0056] The following will introduce in detail the specific implementation solutions provided by the embodiments of the present application.

[0057] Example 1

[0058] First, from the perspective of the server, this Example 1 provides a method for providing product object information. Refer to Figure 2 , this method may include:

[0059] S201: Determine at least one target product object to be provided to the target user and its original description information.

[0060] In the embodiments of the present application, the conversion of the localized expression of the product description information can be performed in a variety of different scenarios. For example, one of the scenarios can be the product recommendation scenario, that is, when it is necessary to recommend products to the target user, at least one target product object to be recommended can be determined according to the historical behavior records, preferences and other information of the target user (this process can already achieve "one product for one person" at the product dimension), and then when specifically displaying the information of this recommended product object on the page, the processing based on the localized expression can also be performed to achieve "one product for one person" at the level of product information expression; or, in the search scenario, when the user inputs keywords or pictures, etc., at least one target product object that meets the search conditions can be matched. At this time, in this search scenario, the prior art can also achieve "one product for one person" at the product dimension, and then when specifically displaying the information of the search results of this product object on the page, the processing based on the localized expression can also be performed to achieve "one product for one person" at the level of product information expression; in addition, when displaying the detailed information page of a specific product to the user, the conversion process of the specific description information based on the localized expression can also be performed, and so on. That is to say, in the embodiments of the present application, the specific conversion process of the localized expression can be applied in a variety of scenarios. Correspondingly, "one product for one person" can be achieved in the foreground expression dimension of the product description information on a variety of different types of pages such as the recommended product information flow page, the product search result page, and the product detailed information page.

[0061] Among them, the original description information of the target product object can be obtained from the product information library, and it is usually the information provided by the merchant when publishing the product, including the title, rich media information (including pictures, videos, audio, etc.), attribute / parameter information, etc. In addition, the original description information of the product can also include the description information such as the user comments obtained by the product, and so on. In short, all the information that needs to be displayed on the specific page can be used as the description information to be converted. When specifically performing the conversion process, all the description information can be converted, or only a part of it can be converted, and so on.

[0062] It should be noted that in actual applications, the same target commodity object may be associated with multiple different SKUs (Stock Keeping Unit, the basic unit for measuring inventory in and out). Although they may share the same detailed page, usually corresponding description information such as pictures and titles will be provided for each SKU respectively. For example, for a clothing item, there are three size rules: S, M, and L, and two colors: black and white. Before placing an order, users usually need to first open the SKU selection interface through the operation options in the detailed page (usually shown in the form of a half-screen floating layer of the detailed page), and select specific sizes, colors, etc. For example, if a user selects size S and color white, it means the user needs to purchase the SKU with size S and color white, and so on. Among them, when displaying various selectable SKU information through the SKU selection interface, different representative pictures may also be provided for different SKUs. For example, in the aforementioned example, different colored clothes can correspond to different pictures, and so on. In the traditional method, when different users view this SKU selection interface, the displayed information is the same, including the pictures corresponding to the same SKU will be the same, and so on. However, in the embodiments of this application, for the description information displayed in this SKU selection interface, localization expression processing based on the country / region to which the current user belongs can also be performed, so that when users in different countries / regions view the description information of the same SKU of the same commodity, it can be different because it has been processed according to the localization expression methods of different countries / regions. Specifically, for the same SKU, the solution of this application considers factors such as country and user information, and uses methods such as AI automatic generation to perform different localization expression processing or personalized processing, including: converting the keywords related to the commodity name and / or adjectives included in the original text content into the commonly used local words corresponding to the country / region attribute information. Or, according to the attribute preference information of the category of the target commodity object by the people corresponding to the country / region attribute information, perform conversion processing on the text content related to the commodity attribute expression in the original title text content. Or, according to the country / region attribute information of the target user, reorder the original user comment text content so as to give priority to displaying the user comment text content localized corresponding to the country / region attribute information, and so on.

[0063] In summary, in the embodiments of the present application, the original description information corresponding to each SKU can also be obtained. Subsequently, during the localization process, the original description information corresponding to such SKUs can also be processed based on the localized expression, so that the target description information about the same SKU seen by users in different countries / regions can be different. For example, the title, pictures, etc. of the SKU can be processed for localized expression, and so on. Specifically, for example, the clothing products in the foregoing example have two colors, black and white, which respectively correspond to different SKU pictures. In the embodiments of the present application, if users A and B are in different countries / regions, then even if both users A and B view the "white" of the product, the SKU pictures displayed corresponding to this "white" can also be different, each having its own different localized expression, and so on.

[0064] S202: Determine the national / regional attribute information of the target user.

[0065] In addition to determining the original description information of the commodity object, the national / regional attribute information of the specific target user can also be determined, including the country / region to which the specific target user belongs, etc. Among them, the national / regional attribute information of the target user can be determined in various ways. For example, the information such as the user's IP (Internet Protocol) address and the common delivery address can be used to determine the information such as the country / region to which the user belongs. Of course, there are also some users who may live in a certain country / region but are actually people from other countries / regions. In this case, the user may still retain the habits of the original country / region, and so on. Therefore, the actual national / regional attribute of the user can also be determined in combination with the user's historical behavior records, etc. For example, if the IP address of a certain user shows that he is located in country A, but based on information such as his historical browsing habits and shopping habits, it can be judged that this user is more likely to be originally from country B. In this case, country B can also be used as the value of the national / regional attribute of this user, and so on.

[0066] It should be noted here that in the specific implementation, in addition to obtaining the national / regional attribute information of the user, other information about the user portrait data can also be obtained, including whether he is male / female, young / middle-aged / old, and so on. These user portrait data can also be used to personalize the front-end description information of the commodity from other dimensions.

[0067] S203: Perform relevant processing for localized expression on the original description information according to the national / regional attribute information of the target user to generate target description information.

[0068] After determining the at least one target commodity object and its original description information, as well as the nationalization / regionalization attribute information of the target user, relevant processing for localizing the expression of the original description information can be performed according to the nationalization / regionalization attribute information of such a target user to generate target description information.

[0069] Among them, the specific original description information can include various different modalities of information, including text information, image information, attribute / parameter description information, etc. Different modalities of information can be processed separately, or, alternatively, some modalities of information can be processed.

[0070] Specifically, for text-based information, it can specifically include the title of the commodity, the graphic and text details content in the detail page, user comment content, etc. All of these may involve text-based information, and thus, all can be subjected to conversion processing. Of course, for pages of commodity list types (such as, for example, the recommended commodity information flow page, or the search result page, etc.), there may be no involvement of graphic and text details, user comments, etc. Therefore, only the title part can be converted, and so on.

[0071] Specifically when performing conversion processing on text information, that is, according to the nationalization / regionalization attribute information of the target user, conversion processing for localizing the expression of the original text information of the target commodity object is performed to generate target text for displaying the target text in the target page.

[0072] Among them, when specifically performing conversion based on localizing the expression of text information, the most basic conversion can include: converting keywords included in the original text information that are related to the commodity name and / or adjectives into local common words corresponding to the nationalization / regionalization attribute information. For example, assuming that the original title of a certain commodity includes "orange", and the current target user is a user from Region C, then the "orange" in the title can be converted to "liuting"; assuming that the original title of a certain commodity includes the adjective "beautiful", and the current target user is a user from Region B, then the "beautiful" in the title can be converted to "liang", and so on.

[0073] In addition, during the process of converting the original title text information, there may also be a situation where the title text may include text content related to the expression of product attributes. At this time, in addition to converting keywords such as nouns and adjectives included therein into commonly used local words, it is also possible to convert the text content related to the expression of product attributes in the original title text information according to the attribute preference information of the products in the category to which the target product object belongs for the population corresponding to the country / region attribute information. Among them, when specifically converting the text content related to the expression of product attributes, there can be various methods. For example, the target text content related to the expression of product attributes included in the original title text information can be deleted or converted into other attribute values, or text content related to the expression of product attributes can be added to the original title text, and so on.

[0074] For example, for products in the category of "thermal flasks", in order to highlight their selling points in terms of capacity, specific parameter values such as "500 ml" may be expressed in the title text. However, for users in some regions, they may be more recognized of the "thermal flasks" with a capacity of "375 ml". At this time, if the product also has the SKU (Stock Keeping Unit, the basic unit for measuring inventory in and out) corresponding to "375 ml", then when displaying the product information to users in this region, the attribute value of "500 ml" originally expressed in the title can be modified to "375 ml". Or, for the product of "power bank", users in Region A may be more concerned about its selling point in terms of "large capacity", so the title of the product includes attribute values such as "20,000 mAh". However, users in Region B may not be too concerned about the capacity information. Therefore, when the target users are users in Region B, the "20,000 mAh" in the title can be deleted, and so on.

[0075] In addition, the specific text - type information may further include user review text content. Specifically, when performing conversion processing, in addition to converting keywords such as nouns and adjectives included therein into commonly used local words, the original user review text content may also be reordered according to the national / regional attribute information of the target user, so as to preferentially display the local user review text content corresponding to the national / regional attribute information. That is to say, in the cross - border e - commerce scenario, users who comment on the same product may come from all over the world. At this time, the comment content of users from the same or similar country / region as the current user may be more valuable to the current user. Therefore, through the method of the embodiments of the present application, the local user review text content corresponding to the current target user's country / region can be preferentially displayed. In this way, when the target user views the user comment content of the target product object, the target user can preferentially view this local user comment content.

[0076] It should be noted that specifically in the process of processing text - type information based on local expression, it can be processed in a traditional way. For example, a corresponding relationship thesaurus between commonly used words in different countries / regions is established in advance through collection and statistics. In this way, when performing conversion processing specifically, the replacement of keywords can be completed by querying the thesaurus. However, in this way, the cost is high, and the scale and type of the thesaurus are very limited. For example, if a certain keyword is not stored in the commonly used words corresponding to a certain country / region in the thesaurus, the conversion may not be achieved, and so on. In addition, this simple keyword replacement method may cause the text after replacement to be less smooth in context, or the grammatical expression after replacement may not match the grammatical expression of the current target user's country / region.

[0077] In view of the above situation, in a preferred manner, the ability of large AI models can be utilized to achieve the conversion of text content based on local expressions. Specifically, since some existing large AI models have extremely rich content and knowledge, can understand the "problems" expressed through natural language, and can also create new content according to specific problems. Among them, the newly created text content usually has good performance in terms of context coherence, etc. Therefore, in the preferred embodiment of the present application, the original title text and the national / regional attribute information of the target user can be input into the large-scale parameter model of artificial intelligence AI, so that the large-scale parameter model of AI can perform the conversion process of localizing the expression of the original title text of the target commodity object. In this way, the converted text content can also be further processed by this large-scale parameter model of AI. For example, according to the local expression mode corresponding to the national / regional attribute information, the converted text content can be processed to conform to the local grammar expression habit and / or context coherence processing to generate the target text content, etc.

[0078] Of course, for some general large AI models, when directly using such models to perform the conversion process of localizing the expression of specific text content, it is equivalent to completely using the existing knowledge of the large AI model to generate content. At this time, there may be situations where the generated text content is not precise enough, etc. Therefore, in the specific implementation, in order to make the content produced by the large AI model more applicable to the application scenarios in the embodiments of the present application, the large AI model can also be trained specifically in advance. That is to say, the content included in the large AI model is very rich, but how to effectively extract valuable information for the scenarios described in the embodiments of the present application is relatively important and is the key to enabling the content finally generated by the large AI model to have a high credibility without human intervention.

[0079] To achieve this goal, in one way, specifically during training, some positive and negative samples can be output to the large AI model. These samples can be text content such as some commodity titles, as well as the conversion results corresponding to multiple different countries / regions. By inputting these samples into the large AI model, the large AI model can learn the "knowledge" in the process of specifically converting the text content based on local expressions, etc.

[0080] Of course, in the above manner, a relatively large amount of work may also be generated in the construction of samples, and the training cost will be relatively high. Therefore, in another way, an AI large model can be used in advance to establish a vocabulary that covers richer information, and then this vocabulary can be used as the input information of the model, enabling the AI large model to complete processing such as the conversion of specific text content based on this pre-constructed vocabulary, which can also achieve the purpose of the usability of the content produced by the AI large model to a certain extent. Specifically, positive and negative samples of common words in multiple countries / regions (a small-scale vocabulary, for example, the corresponding relationships such as "video (Region A) - video (Region B)", etc.) can be prepared in advance and input into the AI large model; in addition, some rules can be set, and these rules can be the special requirement information for localization expressions in some countries / regions. For example, Region B needs to output traditional Chinese characters, Region C needs to output Hakka, and so on. In addition, some key categories can be sorted out, and some rules can be set according to these key categories, and so on. Furthermore, when inputting the above positive and negative samples and rules, the AI large model can be required not only to generate data but also to generate generalized results: that is, the AI large model can be required to generalize based on the input small-scale samples and rules, according to the rules and positive and negative samples. For example, if the sample pair "orange (Region A) - liucheng (Region C)" is included in the sample, search for all words related to the "orange" category on the entire network and which countries / regions' common words these words belong to, and so on. In this way, the AI large model can perform generalization based on a limited sample set. Of course, there should also be rules during the generalization process, and this kind of rule belongs to the relevant knowledge information of the scenario based on the embodiments of the present application and is input into the AI large model. Of course, in addition to constructing a more perfect vocabulary for noun-type keywords, similar processing can also be done for adjective-type keywords, enabling the AI large model to also perform conversions based on localization expressions for adjectives.

[0081] In addition, since there is usually a certain collocation relationship between a name and an adjective, that is, not all names and adjectives can be collocated and used together. Therefore, in order to obtain better results, tags can also be added to specific nouns and adjectives according to the matching relationship between nouns and adjectives to specify which can be used for word matching, and so on.

[0082] Furthermore, after having the above samples and rules, the AI large model can also be allowed to verify and score the content it produces. Through continuous optimization training, the purpose of making the content produced in the application scenario of the embodiments of the present application have a certain degree of usability can be achieved.

[0083] In the above manner, since a richer vocabulary is generated offline, when specifically performing conversion processing based on local expressions for text content, the original text content, the country / region attribute information of the current target user, and the above-mentioned offline-generated vocabulary can be input into the large AI model, enabling the large AI model to replace keywords such as commonly used local words based on this vocabulary. Additionally, it can also perform local processing on the grammar expression of the text or handle the coherence of the context, and so on. In this way, since the vocabulary has been generated offline in advance (of course, it can also be updated regularly), the large AI model can obtain the specific target keywords to be replaced by querying in the vocabulary. Then, it only needs to process aspects such as grammar and context coherence. Therefore, the content production efficiency can be further improved, making it more in line with the requirements of real-time. Of course, if there are no commonly used words corresponding to a certain country / region in the vocabulary, the large AI model can also determine them by searching the entire network's knowledge, and so on.

[0084] In a similar manner, sample information such as the attribute preferences and local grammar expression habits corresponding to multiple countries / regions can also be input into the large AI model, enabling the large AI model to also obtain knowledge related to attribute preferences and local grammar expression habits, etc., so that it can perform local conversion on the content related to attribute expressions included in the text content, and can also make the finally generated text content more in line with local expression habits in terms of grammar expression, etc. In this way, when specifically converting the description information of a certain commodity object for a certain target user, the information actually input into the large AI model can include, in addition to the original text content and the country / region attribute information of the target user, a pre-established knowledge base and / or rule information. Among them, the specific knowledge base can include knowledge information such as commonly used local words, attribute preferences, and / or local grammar expression habits corresponding to multiple countries / regions; the specific rule information can include: special requirement information for local expressions in some countries / regions, and so on. Among them, the specific knowledge base can be a knowledge base generated offline by the large AI model, and so on.

[0085] The above has introduced the conversion process of text content. Another type of descriptive information is the attribute / parameter descriptive information of the commodity object, that is, the original descriptive information to be converted can include attribute / parameter descriptive information. For example, for commodities such as clothing, attributes / parameters include size, color, etc.; for commodities such as utensils, attributes / parameters include capacity, color, etc., and so on. For such attribute / parameter information, since it usually involves measurement units, etc., and the commonly used measurement units for the same parameter may be different in different countries / regions. For example, there are significant differences in the commonly used sizes of commodities in different regions, including shoe sizes, clothing sizes, weights, etc. Specifically, for the size of shoes, a size 6 in the United States corresponds to a size 23.5 in China and a size 37 in Europe, and so on. Therefore, the original attribute / parameter descriptive information of the target commodity object can also be converted into a localized expression according to the national / regional attribute information of the target user to generate target attribute / parameter descriptive information for displaying the target attribute / parameter descriptive information on the target page.

[0086] Among them, specifically when converting attribute / parameter descriptive information, on the one hand, the original attribute / parameter descriptive information of the target commodity object can be converted into a standard or unit commonly used locally according to the national / regional attribute information of the target user. For example, for a certain pair of shoes, in the original "size" attribute, it is expressed as European sizes such as 36, 37, 38, etc. When targeting users in the United States, it can be converted into US sizes such as 5, 5.5, 6, etc., and so on.

[0087] On the other hand, if the target commodity object is associated with multiple SKUs corresponding to different attribute values / parameter values, the multiple SKUs can also be reordered so that the SKUs corresponding to the attribute values / parameter values commonly used locally corresponding to the national / regional attribute information are displayed first. For example, for a certain "baby bottle" commodity, its "capacity" attribute includes different capacity values such as 500ml and 375ml, corresponding to different SKUs. By default, the SKU corresponding to "500ml" is displayed first. However, when targeting users in Region B, since "350ml" is the most recognized capacity by local users, the SKU corresponding to "350ml" can be displayed first. Another example is that for some electrical appliances, there are different SKUs supporting multiple voltage values such as "110V" and "220V". Among them, by default, the SKU corresponding to "220V" is displayed first. However, when targeting users in countries / regions such as Japan, the SKU corresponding to "110V" can be displayed first, and so on.

[0088] Among them, when converting and processing the attribute / parameter description information of a commodity based on local expression, it can also be achieved through traditional methods. Alternatively, the original attribute / parameter description information and the national / regional attribute information of the target user can be input into the large-scale AI parameter model, so that the large-scale AI parameter model can perform conversion processing on the original attribute / parameter description information of the target commodity object for local expression. Of course, in order to enable the large-scale AI model to perform more accurate conversion processing on the description information related to attributes / parameters, the large-scale AI model can also be trained specifically in advance. Or, the common expression methods of attributes in different dimensions in different countries / regions, etc., can be input into the large-scale AI model as knowledge-based information, so that the large-scale AI model can output more usable content, and so on.

[0089] In addition to text content and attribute / parameter description information, the specific commodity description information can also include rich media information, such as image and audio information, and so on. That is to say, the original description information that specifically needs to be converted can also include the rich media information of the commodity object, which can specifically include pictures, videos, audio, and so on. Therefore, the original rich media information of the target commodity object can also be converted for local expression according to the national / regional attribute information of the target user to generate target rich media information, so as to display the target image information on the target page.

[0090] Among them, when specifically processing image information, it can include conversion processing of the composition style, model type, and / or atmosphere elements of the image for local expression. Among them, the commodity composition can be regenerated according to local preferences. For example, users in Brazil tend to prefer colorful compositions, while users in Europe and America prefer more minimalist styles. Therefore, the composition style of the image can be converted according to the preferences of users in different countries / regions. Regarding the model type, there are significant differences in the body shapes of users in Europe, America, and Asia. Asian users may be more suitable for models with a fresh style and a slender figure, while users in Europe and America prefer more voluptuous models. In addition, since users in Europe and America generally have a more voluptuous body shape, using a matching model for display will give users a more immersive experience. Regarding the atmosphere elements in the image, they are usually elements added to set off a certain atmosphere. However, some of these elements may be sensitive elements for users in certain countries / regions. For example, for African users, especially African-American users in the United States, "watermelon" may be a sensitive element, and so on. Therefore, such atmosphere elements can also be converted, such as deleting some sensitive elements or replacing them with other elements, and so on.

[0091] Specifically, when the image content is converted based on localized expression, it can also be completed through AI large-scale parameter models. In specific implementation, the AI ​​large model can perform subject recognition, cutout, replacement and other processing on the specific image. In particular, for the process of model character type conversion, the AI ​​large model can generate the corresponding model character image according to the country / region to which the target user belongs, and re-match it with the product image of the current product to generate a new model upper body image. For example, in an image, it was originally an Asian model wearing a certain skirt. When showing the product to users in Europe and the United States, it can be replaced with a model character with the general appearance and figure of European and American users, and the picture of the skirt can be matched to the new model character after deformation processing. Then, it is synthesized with specific background images, atmosphere elements, etc. into the final target image, and so on.

[0092] It should be noted here that the converted model characters can be virtual model characters created by the AI ​​big model. In order to improve processing efficiency, virtual model characters of some representative countries / regions can be generated in advance in an offline manner. In this way, when performing the conversion, one of the virtual model characters can be selected from the pre-saved virtual model characters, and then the product image of the specific product can be matched to the virtual model character image, etc. Among them, the same country / region can correspond to multiple different virtual model characters. In this way, for a page including multiple products, different virtual model characters can be used to match different product images to avoid the situation where multiple products on the page have the same virtual model character image.

[0093] It should also be noted that when matching the product image with the virtual model character image, the specific attributes / parameters of the product can also be taken into consideration. For example, the fabric of some clothing may not be elastic, or the size may be small, which may not be suitable for European and American users with plump bodies. Therefore, when showing it to users in Europe and the United States, it is not necessary to replace it with the corresponding model character image in Europe and the United States to avoid misleading consumers. In specific implementation, if the model character is converted through the AI ​​big model, the attribute / parameter information of the specific product can also be input into the AI ​​big model, so that the AI ​​big model can output the specific conversion result through comprehensive decision-making.

[0094] If the original description information of a specific product object also includes audio information, such as background music of a video, then such audio information can also be converted based on localized expressions. For example, if the target user is a user in India, then the background music of the specific product can be replaced with audio with an Indian style, and so on.

[0095] S204: Provide the target description information corresponding to the at least one target commodity object to the client corresponding to the target user, so as to provide the target description information through the target page.

[0096] After performing conversion processing based on local expression on the text content, attribute / parameter description information, rich media information, etc. of the target commodity object, it can be returned to the client corresponding to the target user, and the client provides it to the target user through a specific target page. Among them, the specific target page can be a recommended commodity information flow page, or a commodity search result page, or a commodity detail page, etc.

[0097] In summary, through the embodiments of the present application, when it is necessary to provide information about a target commodity object to a target user, the original description information of such a target commodity object and the national / regional attribute information of the target user can be obtained first. Then, according to the national / regional attribute information of the target user, relevant processing for local expression can be performed on the original description information to generate target description information, so as to provide such target description information to the target user through the target page. In this way, it can be achieved that the expression of the foreground description information of the commodity object provided in the target page is "one size fits one person", and it is more in line with the local preferences of the country / region where the current target user belongs, thereby more easily stimulating the user's shopping enthusiasm, improving the user experience, and further being beneficial to improving indicators such as click-through rate and conversion rate.

[0098] Among them, in a preferred manner, the capabilities of the AI large-scale parameter model in aspects such as multi-modal content understanding and generation can be utilized to help complete the conversion based on local expression of the commodity description information, which can improve efficiency, and can also improve the context coherence of the converted text content, etc. Of course, in order to make the content generated by the AI large model have higher usability or accuracy, some samples, rules, etc. can also be used to perform targeted training on the AI large model in advance.

[0099] Embodiment 2

[0100] This Embodiment 2 corresponds to Embodiment 1. From the perspective of the client, a method for providing commodity object information is provided. See Figure 3 , and this method may specifically include:

[0101] S301: Receive the target description information of at least one target commodity object provided by the server for the target user, where the target description information is generated by performing relevant processing for local expression on the original description information of the target commodity object according to the national / regional attribute information of the target user;

[0102] S302: Display the target description information of the at least one target commodity object through the target page.

[0103] Regarding the parts not described in detail in the second embodiment, reference may be made to the foregoing first embodiment or the descriptions in other parts of this specification, which will not be elaborated here.

[0104] It should be noted that the embodiments of this application may involve the use of user data. In practical applications, user-specific personal data can be used in the solutions described herein within the scope permitted by applicable laws and regulations (for example, with the user's explicit consent, giving the user a practical notice, etc.) and in compliance with the requirements of the applicable laws and regulations of the country where it is located.

[0105] Corresponding to the first embodiment, the embodiments of this application also provide a device for providing commodity object information. Refer to Figure 4 , and the device may include:

[0106] An original description information determination unit 401, configured to determine at least one target commodity object to be provided to a target user and its original description information;

[0107] A user attribute information determination unit 402, configured to determine the country / region-specific attribute information of the target user;

[0108] A localization processing unit 403, configured to perform relevant processing based on localization expression on the original description information according to the country / region-specific attribute information of the target user to generate target description information;

[0109] An information providing unit 404, configured to provide the target description information corresponding to the at least one target commodity object to the client corresponding to the target user, so as to provide the target description information through the target page.

[0110] Specifically, the localization processing unit may specifically be used for:

[0111] Use an artificial intelligence AI large-scale parameter model to perform model understanding on the original description information, and perform relevant processing based on localization expression on the original description information according to the country / region-specific attribute information of the target user to generate the target description information.

[0112] Wherein, the original description information includes: the original text content of the target commodity object;

[0113] At this time, the localization processing unit may specifically be used for:

[0114] According to the national / regional attribute information of the target user, perform a conversion process of localizing the original text content of the target commodity object to generate the target text content, so as to display the target text content on the target page.

[0115] Specifically, keywords related to the commodity name and / or adjectives included in the original text content can be converted into commonly used local words corresponding to the national / regional attribute information.

[0116] Among them, the original text content includes: the original title text content;

[0117] At this time, the localization processing unit can specifically also be used for:

[0118] According to the attribute preference information of the commodities in the category to which the target commodity object belongs for the population corresponding to the national / regional attribute information, perform a conversion process of the text content related to the commodity attribute expression on the original title text content.

[0119] Among them, specifically, the target text content related to the commodity attribute expression included in the original title text content can be deleted or converted into other attribute values, or text content related to the commodity attribute expression can be added to the original title text content.

[0120] Or, the original text content includes: the original user review text content;

[0121] At this time, the localization processing unit can specifically also be used for:

[0122] According to the national / regional attribute information of the target user, reorder the original user review text content so as to preferentially display the user review text content that belongs to the localization corresponding to the national / regional attribute information.

[0123] Specifically, the localization processing unit can specifically be used for:

[0124] Input the original text content and the national / regional attribute information of the target user into the AI large-scale parameter model, so that the AI large-scale parameter model performs a conversion process of localizing the original text content of the target commodity object;

[0125] The AI large-scale parameter model is also used to perform processing that conforms to the local grammar expression habit and / or context coherence processing on the text content after the conversion process according to the local expression mode corresponding to the national / regional attribute information, so as to generate the target text content.

[0126] Among them, the information input into the AI large-scale parameter model may further include: a pre-established knowledge base and / or rule information. The knowledge base includes knowledge information such as localized common words, attribute preference information, and / or localized grammar expression habits corresponding to multiple countries / regions; the rule information includes: special requirement information for localized expressions in some countries / regions.

[0127] In addition, the original description information includes: the original attribute / parameter description information of the target commodity object;

[0128] At this time, the localization processing unit can specifically be used for:

[0129] According to the country / region attribute information of the target user, perform a conversion process of localized expression on the original attribute / parameter description information of the target commodity object to generate target attribute / parameter description information, so as to display the target attribute / parameter description information on the target page.

[0130] Specifically, according to the country / region attribute information of the target user, convert the original attribute / parameter description information of the target commodity object into a standard or unit commonly used in localization.

[0131] Or, if the target commodity object is associated with multiple stock-keeping units (SKUs) with different attribute values / parameter values, reorder the multiple SKUs so as to preferentially display the SKUs corresponding to the attribute values / parameter values commonly used in localization corresponding to the country / region attribute information.

[0132] Specifically, the localization processing unit can specifically be used for:

[0133] Input the original attribute / parameter description information and the country / region attribute information of the target user into the AI large-scale parameter model, so that the AI large-scale parameter model performs a conversion process of localized expression on the original attribute / parameter description information of the target commodity object.

[0134] In addition, the original description information includes: the original rich media information of the target commodity object;

[0135] At this time, the localization processing unit can specifically be used for:

[0136] According to the country / region attribute information of the target user, perform a conversion process of localized expression on the original rich media information of the target commodity object to generate target rich media information, so as to display the target rich media information on the target page.

[0137] Specifically, the original rich media information includes original image information;

[0138] At this time, the localization processing unit may specifically be configured to:

[0139] According to the national / regional attribute information of the target user, perform conversion processing on the composition style, model character type, and / or atmosphere elements of the original image information to generate target image information that conforms to the localization preferences corresponding to the national / regional attribute information.

[0140] Specifically, the original image information and the national / regional attribute information of the target user may be input into the AI large-scale parameter model, so that the AI large-scale parameter model performs conversion processing on the composition style, model character type, and / or atmosphere elements of the original image information.

[0141] Among them, when performing conversion processing on the model character type in the original image information, it is possible to determine whether to perform conversion processing on the model character type according to the attribute / parameter information of the target commodity object.

[0142] In addition, the original rich media information may include original audio information;

[0143] At this time, the localization processing unit may specifically be configured to:

[0144] According to the national / regional attribute information of the target user, perform conversion processing on the original audio information to generate target audio information that conforms to the localization preferences corresponding to the national / regional attribute information.

[0145] Corresponding to Embodiment 2, an embodiment of the present application further provides a device for providing commodity object information. Refer to Figure 5 , the device may include:

[0146] An information receiving unit 501, configured to receive target description information of at least one target commodity object provided by the server for the target user, where the target description information is generated after performing relevant processing based on localization expression on the original description information of the target commodity object according to the national / regional attribute information of the target user;

[0147] An information providing unit 502, configured to provide the target description information of the at least one target commodity object through a target page.

[0148] In addition, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of the method described in any one of the foregoing method embodiments are implemented.

[0149] And an electronic device, including:

[0150] One or more processors; and

[0151] A memory associated with the one or more processors, the memory being configured to store program instructions that, when read and executed by the one or more processors, perform the steps of the method according to any one of the foregoing method embodiments.

[0152] Wherein Figure 6 Exemplarily shows the architecture of an electronic device. For example, device 600 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, an aircraft, etc.

[0153] Referring to Figure 6 , device 600 may include one or more of the following components: a processing component 602, a memory 604, a power supply component 606, a multimedia component 608, an audio component 610, an input / output (I / O) interface 612, a sensor component 614, and a communication component 616.

[0154] The processing component 602 generally controls the overall operation of device 600, such as operations associated with display, telephone calls, data communication, camera operations, and recording operations. The processing component 602 may include one or more processors 620 to execute instructions to complete all or part of the steps of the method provided by the technical solution of the present disclosure. In addition, the processing component 602 may include one or more modules to facilitate the interaction between the processing component 602 and other components. For example, the processing component 602 may include a multimedia module to facilitate the interaction between the multimedia component 608 and the processing component 602.

[0155] The memory 604 is configured to store various types of data to support the operation of device 600. Examples of such data include instructions for any application or method operating on device 600, contact data, phone book data, messages, pictures, videos, etc. The memory 604 may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, a magnetic disk, or an optical disc.

[0156] The power supply component 606 provides power to various components of device 600. The power supply component 606 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for device 600.

[0157] The multimedia component 608 includes a screen that provides an output interface between the device 600 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen can be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors can sense not only the boundaries of touch or swipe actions but also detect the duration and pressure associated with the touch or swipe operation. In some embodiments, the multimedia component 608 includes a front camera and / or a rear camera. When the device 600 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera can receive external multimedia data. Each of the front camera and the rear camera can be a fixed optical lens system or have a focal length and optical zoom capabilities.

[0158] The audio component 610 is configured to output and / or input audio signals. For example, the audio component 610 includes a microphone (MIC) that is configured to receive external audio signals when the device 600 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals can be further stored in the memory 604 or transmitted via the communication component 616. In some embodiments, the audio component 610 further includes a speaker for outputting audio signals.

[0159] The I / O interface 612 provides an interface between the processing component 602 and a peripheral interface module, which can be a keyboard, a click wheel, buttons, etc. These buttons can include but are not limited to: a home button, a volume button, a power button, and a lock button.

[0160] The sensor component 614 includes one or more sensors for providing an assessment of the various aspects of the state of the device 600. For example, the sensor component 614 can detect the on / off state of the device 600, the relative positioning of components, such as the display and the keypad of the device 600. The sensor component 614 can also detect a change in the position of the device 600 or a component of the device 600, the presence or absence of user contact with the device 600, the orientation or acceleration / deceleration of the device 600, and the temperature change of the device 600. The sensor component 614 can include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor component 614 can also include a light sensor, such as a CMOS or a CCD image sensor, for use in imaging applications. In some embodiments, the sensor component 614 can further include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0161] The communication component 616 is configured to facilitate communication between the device 600 and other devices in a wired or wireless manner. The device 600 can access a communication standard-based wireless network, such as WiFi, or a mobile communication network such as 2G, 3G, 4G / LTE, 5G, etc. In an exemplary embodiment, the communication component 616 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 616 further includes a Near Field Communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on Radio Frequency Identification (RFID) technology, Infrared Data Association (IrDA) technology, Ultra-Wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0162] In an exemplary embodiment, the device 600 can be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components for performing the above method.

[0163] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 604 including instructions that can be executed by the processor 620 of the device 600 to complete the method provided by the technical solution of the present disclosure. For example, the non-transitory computer-readable storage medium can be a ROM, Random Access Memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0164] From the description of the above embodiments, those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disc, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present application.

[0165] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other, and the key point of each embodiment is to illustrate the differences from other embodiments. In particular, for a system or a system embodiment, since it is basically similar to a method embodiment, the description is relatively simple, and for the relevant parts, reference can be made to the description of the method embodiment. The systems and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative efforts.

[0166] The method and electronic device for providing commodity object information provided by this application have been introduced in detail above. Specific examples are used herein to elaborate on the principle and implementation manner of this application. The description of the above embodiments is only used to help understand the method and its core idea of this application; at the same time, for those of ordinary skill in the art, there will be changes in the specific implementation manner and application scope according to the idea of this application. In summary, the content of this specification should not be construed as a limitation to this application.

Claims

1. A method for providing information on a commodity object, characterized in that, Including: Determine at least one target commodity object to be provided to a target user through a target page and its original description information; Determine the country / region localization attribute information of the target user; According to the country / region localization attribute information of the target user, perform conversion processing related to local expression preferences on the original description information to generate target description information; wherein, the original description information includes: the original title text content of the target commodity object, and the conversion processing related to local expression preferences includes: according to the attribute preference information of the commodities in the category to which the target commodity object belongs for the population corresponding to the country / region localization attribute information, perform conversion processing on the text content related to commodity attribute expression in the original title text; wherein, the target commodity object is associated with multiple stock - keeping units (SKUs), and different SKUs correspond to different specification parameter values, and the text content related to commodity attribute expression in the original title text includes: the specification parameter value content corresponding to one of the SKUs included in the original title text, and the conversion processing includes: according to the preference information of the specification parameter values of the commodities in the category to which the target commodity object belongs for the population corresponding to the country / region localization attribute information, convert the specification parameter value content in the original title text into the corresponding specification parameter value content of other SKUs; Provide the target description information corresponding to the at least one target commodity object to the client corresponding to the target user, so as to provide the target description information through the target page, so that when providing information about the same commodity object to users with different country / region localization attribute information through the target page, different target description information generated based on different local expression preferences is provided.

2. The method according to claim 1, wherein: The conversion processing related to local expression preferences on the original description information further includes: Convert the keywords related to commodity names and / or adjectives included in the original title text content into local common words corresponding to the country / region localization attribute information.

3. The method according to claim 1, wherein: The related processing of local expression of the original description information to generate target description information further includes: Use an artificial intelligence (AI) large - scale parameter model to perform processing that conforms to local grammar expression habits and / or context coherence processing on the title text content after local expression conversion to generate the target description information.

4. A method for providing commodity object information, characterized in that, Including: Determine a target commodity object to be provided to a target user through a target page and its original description information, and the target page includes the commodity details page of the target commodity object; Determine the country / region localization attribute information of the target user; Based on the country / region localization attribute information of the target user, perform transformation processing related to localization expression preferences on the original description information to generate target description information; wherein, the transformation processing related to localization expression preferences includes: if the target commodity object is associated with multiple SKUs corresponding to different attribute values / specification parameter values, when displaying the SKU selection interface through the commodity details page, reorder the multiple SKUs so as to preferentially display the SKUs corresponding to the attribute values / parameter values commonly used in localization corresponding to the country / region localization attribute information.

5. A method for providing product object information, characterized in that, including: Determine at least one target commodity object to be provided to the target user through the target page and its original description information; Determine the country / region localization attribute information of the target user; Based on the country / region localization attribute information of the target user, perform transformation processing related to localization expression preferences on the original description information to generate target description information; wherein, the original description information includes: the original rich media information of the target commodity object; the transformation processing related to localization expression preferences includes: using an AI large-scale parameter model, based on the country / region localization attribute information of the target user, perform transformation processing on the original rich media information of the target commodity object for localization expression to generate target rich media information, so as to display the target rich media information on the target page; wherein, the original rich media information includes original image information and / or original audio information, and the transformation processing includes: based on the country / region localization attribute information of the target user, perform transformation processing on the composition style, model type, and / or atmosphere elements of the original image information to generate target image information that conforms to the localization preferences corresponding to the country / region localization attribute information, and / or, based on the country / region localization attribute information of the target user, perform transformation processing on the original audio information to generate target audio information that conforms to the localization preferences corresponding to the country / region localization attribute information.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

7. An electronic device, characterized in that, including: One or more processors; and A memory associated with the one or more processors, the memory is used to store program instructions, and when the program instructions are read and executed by the one or more processors, they execute the steps of the method according to any one of claims 1 to 5.

Images