Data object putting method and electronic equipment

Abstract

The embodiment of the invention discloses a data object putting method and electronic equipment, and the method comprises the steps: determining a range of a candidate second data object for a first data object according to following putting target information configured by a first user for the first data object; determining a target second data object if at least one second data object exists in the determined to-be-displayed data object set in the process of recalling and sorting the data objects based on a data object search request of a second user or a demand of recommending the data objects to the second user; and performing dynamic bidding on the first data object according to the sorting position where the target second data object is located so as to display the first data object near a resource niche where the target second data object is located in a data object set display page. Through the embodiment of the invention, the data object putting effect can be improved.

Description

Technical Field

[0001] This application relates to the field of information processing technology, and in particular to methods and electronic devices for data object delivery. Background Technology

[0002] Product promotion typically refers to advertising selected products to increase their exposure and visibility to consumers. For example, ads can be targeted to consumers based on whether their search keywords are related to the promoted product, making it easier for them to discover the product and potentially leading to a purchase. Alternatively, ads can be placed within product recommendation scenarios, and so on.

[0003] Traditionally, merchants typically need to select products to promote and then drive traffic to their stores through keyword advertising. If a merchant advertises a specific keyword for a promoted product, that product has a chance to appear in search results when consumers search for that keyword or related product attributes. However, since different merchants may compete for the same or similar products, and different merchants might choose to advertise the same or similar products using the same keywords, merchants need to bid on those specific keywords. This allows for ranking different products based on their bids when similar products from different merchants hit the same keywords. In short, current product promotion processes require merchants to select keywords / target audiences relevant to the promoted products, bid on those keywords / target audiences, and continuously iterate their strategies. These processes usually require merchants to have strong advertising operation experience.

[0004] In addition, some systems can provide merchants with intelligent advertising services. That is, merchants only need to specify the products they want to promote, and then the system can automatically place some keywords (since multiple keywords are usually placed for each product, these multiple keywords can be called "keyword packages") or target audiences for the specific products and automatically bid. The process of automatically placing keywords / target audiences and bidding can also be constantly changing. For example, the keyword package can be updated for specific products every day, and so on.

[0005] This intelligent ad delivery service can help merchants improve efficiency and reduce the need for extensive advertising experience. However, for some products, it can be difficult to determine an accurate ad delivery strategy, and the effectiveness of the campaign is often hard to guarantee. For example, for newly launched products, the lack of user data may make it difficult to reach the target audience, and existing product ad delivery systems cannot yet provide a solution to this problem. Summary of the Invention

[0006] This application provides a data object delivery method and electronic device, which helps to improve the effectiveness of data object delivery.

[0007] This application provides the following solution:

[0008] A method for delivering data objects, comprising:

[0009] Based on the follow-up investment target information configured by the first user for the first data object, determine the range of candidate second data objects for the first data object;

[0010] In the process of recalling and sorting data objects based on the data object search request of the second user or the need to recommend data objects to the second user, if at least one second data object exists in the determined set of data objects to be displayed, then the target second data object is determined.

[0011] The first data object is dynamically bid on based on its sorting position, so that it is displayed near the resource position of the target second data object on the data object collection display page.

[0012] The target information for follow-up investment includes: data object feature information set according to the category to which the first data object belongs;

[0013] Determining the range of candidate second data objects for the first data object includes:

[0014] Based on the data objects in the data object information database that match the feature information, the range of candidate second data objects is selected.

[0015] This also includes:

[0016] Multiple optional feature tags are provided so that the first user can express the feature information of the data object and determine the follow-up investment target information by selecting feature tags; wherein, the feature tags are defined in advance after classifying the data object follow-up investment targets in the follow-up investment scenario; the feature tags are associated with corresponding data object selection rule information.

[0017] The feature tags include: similar data objects, which are used to set the follow-up target information as follows: follow data objects under the leaf category to which the first data object belongs, whose similarity to the first data object meets the conditions, and whose prices are similar.

[0018] The feature tags include: high-sales data objects, which are used to set the follow-up target information to: follow data objects under the leaf category to which the first data object belongs, whose sales volume and / or sales growth rate meet the conditions for deployment.

[0019] The feature tags include: matching data objects, used to set the follow-up target information as follows: to follow data objects with similar purchasing intentions under the matching category of the first data object for targeting.

[0020] The step of determining the range of candidate second data objects for the first data object based on the follow-up target information configured by the first user for the first data object includes:

[0021] An AI model based on collaborative filtering algorithm is used to mine information from user association behavior data across categories to identify data objects with similar purchase intentions under the matching categories of the first data object, and these data objects are selected as the candidate second data objects.

[0022] This also includes:

[0023] Operational options are provided for configuring delivery preferences to narrow down the range of candidate second data objects based on the corresponding delivery preference information.

[0024] The target information for follow-up investment includes: identification information of at least one specified data object configured by the first user;

[0025] Determining the range of candidate second data objects for the first data object includes:

[0026] At least one specified data object configured by the first user is identified as a candidate second data object.

[0027] The specified data object includes: other data objects published by the first user that have the same or similar sales attributes as the first data object.

[0028] Wherein, if the first data object is a newly published data object by the first user, the specified data object includes: a data object published in the past by the same first user, or a data object published in the past by other first users that has the same or similar sales attributes as the first data object.

[0029] The target information for follow-up investment includes: brand name or store name information configured by the first user;

[0030] Determining the range of candidate second data objects for the first data object includes:

[0031] Data objects associated with the brand name or store name and belonging to the same leaf category as the first data object are identified as candidate second data objects.

[0032] If the sorting includes a coarse sorting stage and a fine sorting stage, then in the fine sorting stage, it is determined whether there is at least one second data object in the set of data objects to be displayed, and a dynamic bid is made for the first data object.

[0033] If, during the data object recall process, the first data object is not recalled by the recall algorithm, and there is at least one second data object in the set of data objects to be displayed, then the first data object is added to the set of data objects to be displayed, and a dynamic bid is made for the first data object so that the first data object is displayed near the resource position where the target second data object is located.

[0034] A computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the steps of any of the preceding methods.

[0035] An electronic device, comprising:

[0036] One or more processors; and

[0037] A memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the steps of the method described in any of the preceding descriptions.

[0038] A computer program product includes a computer program / computer executable instructions that, when executed by a processor in an electronic device, implement the steps of any of the preceding methods.

[0039] According to the specific embodiments provided in this application, the following technical effects are disclosed:

[0040] This application provides a "follow-up" intelligent ad placement method for merchants and other first users. In this method, the first user can configure follow-up target information for a first data object. The system can then determine a range of candidate second data objects based on the follow-up target information configured by the first user. Subsequently, during the process of recalling and sorting data objects based on the second user's data object search request or the need to recommend data objects to the second user, if at least one second data object exists in the determined set of data objects to be displayed, the target second data object can be identified first. Then, the first data object is dynamically bid on based on the ranking position of the target second data object, so that the first data object is displayed near the resource position of the target second data object on the data object set display page. In this way, for data objects where it is difficult to determine a clear ad placement strategy or achieve good ad placement results from the data object's own perspective, following the second data object can be chosen, allowing the first data object to have the opportunity to gain exposure and conversion through the target second data object, thereby improving ad placement results. Of course, even if the data object itself does not have problems such as difficulty in determining an ad placement strategy, the solution provided in this application embodiment can still be used for follow-up ad placement.

[0041] Of course, any product implementing this application does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a schematic diagram of the system architecture provided in the embodiments of this application;

[0044] Figure 2 This is a flowchart of the method provided in the embodiments of this application;

[0045] Figure 3 This is a schematic diagram of the first interface provided in an embodiment of this application;

[0046] Figure 4 This is a schematic diagram of the second interface provided in an embodiment of this application;

[0047] Figure 5 This is a schematic diagram of the third interface provided in an embodiment of this application;

[0048] Figure 6This is a schematic diagram of the electronic device provided in the embodiments of this application. Detailed Implementation

[0049] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0050] To facilitate understanding of the solutions provided in this application, it is important to first clarify that in traditional advertising-based promotion of data objects (specifically, product objects in a product information service system; for ease of description, this application primarily uses products as an example), whether merchants manually select keywords / target audiences and set bidding strategies, or the system algorithm automatically selects keywords / target audiences and sets bids in a smart advertising approach, the promotion is primarily from the perspective of the product itself. Specifically, when setting bids, the estimated conversion rate and other metrics of the product itself are mainly considered. After recalling a batch of products in a product search or recommendation scenario, the ranking position of different products can be determined based on the estimated conversion rate of the specific recalled products in the corresponding keywords / target audiences (how likely a user would purchase the product if it were displayed in search results or recommendations), bidding, etc. Generally, with the same estimated conversion rate, products with higher bids have a greater chance of being displayed first. However, for some products, accurate estimated conversion rate information may be difficult to obtain, making it difficult to achieve effective advertising results. For example, for some new products, due to a lack of data support, information such as the target high-quality audience has not yet been effectively obtained. At this time, it is difficult to predict the conversion rate and other data that the new product may obtain when it is displayed to a certain consumer user. Consequently, the effectiveness of the campaign is difficult to guarantee.

[0051] In this embodiment, a "follow-up" scheme is provided. This means that for a product that needs to be promoted through advertising (for ease of description, this embodiment refers to a product with advertising needs as the "first product"), at least one candidate second product can be selected, and then the advertising strategy of this second product can be followed. In other words, a first product A can be advertised alongside a second product B. Correspondingly, the bidding strategy of the first product A can be related to the display position of the second product B on the product search results page, rather than solely to the conversion rate prediction of the first product A itself. Then, the first product A can be displayed near the resource position of the second product B. That is, consumers can see the first product A near the resource position of the second product B, thereby bringing exposure opportunities to the first product A through the second product B, and potentially leading to corresponding purchase conversions.

[0052] In this approach, merchants no longer need to purchase keywords / target audiences for the first product; instead, they can follow the advertising strategy of the second product, B. Even if the first product isn't recalled by the algorithm, it still has a chance to gain exposure by following the second product's advertising strategy. For example, suppose the first product, A, chooses a follow-the-leader strategy, advertising alongside the second product, B. Here, the second product, B, is a regular advertising item; that is, in the traditional way, the merchant purchases a keyword for the second product, B, or in the smart advertising approach, the system automatically purchases that keyword for the second product, etc. When a consumer searches using that keyword, the second product, B, will be recalled, and its placement within the set of products to be displayed can be determined based on conversion rate estimates and bids for that keyword. At this point, because the merchant may... Since no keywords were purchased for the first product A, there are two possible scenarios: one is that product A, as an organic product, can be recalled; the other is that product A cannot be directly recalled by the recall algorithm. In the first scenario, the bid for product A can be controlled to ensure that product A is displayed near the resource position of product B. In the second scenario, product A can first be added to the set of products to be displayed, and then the bid for product A can be controlled to ensure that product A is displayed near the resource position of product B.

[0053] The follow-up advertising scheme provided in this application can meet the needs of merchants in various scenarios. For example, suppose a merchant's first product A is a new product, which faces problems such as difficulty in acquiring the target high-quality audience and slow sales growth. In this case, the merchant can choose to follow the advertising strategy of the "old products" in its own store as the follow-up target, thereby enabling the new product to acquire the target high-quality audience more quickly. Or, suppose a merchant's first product A may have some competitors in the industry. In this case, the merchant can choose to follow the advertising of these competitors. That is, these competitors can be the second product B to be followed. By following the advertising strategy of the competitors, traffic can be brought to the first product, and so on.

[0054] There are several ways to set up follow-up targets. For example, in one approach, if there are specific brands or products to follow, merchants can specify the products they want to follow through a custom method. Alternatively, if the specific brands or products to follow are not clearly defined, a more fuzzy follow-up approach can be used. In this case, merchants can select the characteristics of the products they want to follow. Specifically, various optional feature tags can be provided, such as follow-up on similar products, follow-up on best-selling products, follow-up on complementary products, and so on. Then, based on the feature tags selected by the merchant, the specific products to follow can be selected.

[0055] From a system architecture perspective, see Figure 1This application embodiment provides a product placement management backend for merchants in the product promotion system. Merchants can use the relevant operation interface of this backend to determine the first product to be promoted and configure the follow-up targets. When configuring the follow-up targets, specific products can be specified for precise follow-up, or specific product feature tags can be selected for fuzzy follow-up, and so on. After the merchant completes the configuration of the follow-up targets for the first product, the product placement server can determine the set of candidate second products. If the merchant has configured precise follow-up, the product specified by the merchant can be directly determined as the candidate second product. If the configured follow-up target is fuzzy follow-up, the range of candidate second products can be selected based on the follow-up target information configured by the merchant, that is, determining which products can be followed by the current first product. The specific selection of second products can be updated periodically, for example, it may be re-selected on a daily basis, etc. After identifying the candidate second product, when a second user initiates a product search or when a product recommendation needs to be made to a second user, the product search / recommendation server will call the product recall and ranking server, which will then perform product recall and ranking processing. During this process, the product delivery server can obtain the set of products to be displayed generated after recall and ranking, and determine whether at least one of the aforementioned candidate second products exists in this set. If so, a target product (the target second product) can be identified, and the first product will be dynamically bid on based on the ranking position of the target second product, ensuring that the first product's ranking position is near the second product. Then, after returning the set of products to be displayed and the ranking results to the product search or recommendation server, the first product can be displayed near the resource position of the target second product on the product search results page or product recommendation page. In this way, the first product can follow the second product in delivery, thereby leveraging the second product's delivery strategy to bring traffic to the first product. For new products, following established products in advertising campaigns can solve the problem of quickly acquiring targeted traffic due to a lack of user data. Alternatively, for potential products (whose absolute sales volume may not be high, but whose sales volume is continuously growing) or best-selling products (products with relatively high absolute sales volume), following competitors in advertising campaigns can help potential products quickly become best-sellers and best-selling products maintain their industry ranking.

[0056] The specific implementation schemes provided in the embodiments of this application will be described in detail below.

[0057] First, this application provides a method for deploying data objects, see [link to relevant documentation]. Figure 2 The method may specifically include the following steps:

[0058] S201: Based on the follow-up target information configured by the first user for the first data object, determine the range of candidate second data objects for the first data object.

[0059] In this system, the first user can be a merchant or seller, while the consumer or buyer is referred to as the second user. The data object can refer to products within the product information service system. The first data object can be the product currently being advertised, and the second data object can be the product that follows it; that is, the first data object can be advertised alongside the second data object. In practice, the first user can specify the first product to be promoted and configure specific follow-up target information through the configuration backend. Then, the system determines the range of candidate second products based on the specific first product and the corresponding follow-up target information.

[0060] In the follow-up bidding scenario, the system automatically bids for the first product, thus constituting a type of smart bidding. However, in practical applications, smart bidding may include various bidding objectives, such as increasing sales volume and stabilizing ROI. The "follow-up bidding" provided in this embodiment can also be considered one such bidding objective. Therefore, in its implementation, it can be displayed to the first user along with other bidding objectives. If the first user selects a bidding objective related to "follow-up bidding," the follow-up bidding process in this embodiment can be triggered. For example... Figure 3 The diagram shows a configuration backend interface provided by the system in one specific implementation. After the user specifies the first product to be promoted, if it is necessary to follow up with other products, the user can select "Smart Bidding" (option 31) in the "Bidding Method" options and "Similar Product Follow-up" (option 32) in the "Bidding Target" options. It should be noted that in practical applications, the second product actually being followed may be a similar product to the first product, or it may be other products, such as popular items, bundled items, etc. Figure 3 The term "follow-up bidding for similar products" is merely a customary term and should not be considered a limitation on the scope of protection of this application.

[0061] After selecting "Similar Product Follow-up," the first user can perform specific operations to configure the follow-up target. Specifically, the first user can configure the follow-up target in several ways, including fuzzy follow-up and precise follow-up. In the fuzzy follow-up mode, product feature information can be set according to the category to which the first product belongs. At this time, the system can select a range of candidate second products based on products in the product information database that match the feature information.

[0062] To facilitate fuzzy follow-up targeting by the first user, multiple selectable feature tags can be provided. This allows the first user to express product characteristic information by selecting feature tags, thereby determining the specific follow-up target information. These feature tags can be pre-defined after classifying the product follow-up targets in the follow-up scenario. That is, in the case of fuzzy follow-up targeting, the first user may have different categories of follow-up targets, such as similar products, best-selling products (high-volume products), and complementary products. Therefore, corresponding feature tags can be provided for the user to choose from. For example, multiple feature tags such as similar products, high-volume products, and complementary products can be provided.

[0063] Specifically, the "similar product" tag can be used to target products within the same leaf category as the first product that meet the similarity criteria and have similar prices. The "high-selling product" tag can be used to target products within the same leaf category as the first product that meet the sales volume and / or sales growth criteria. The "combination product" tag can be used to target products within the same combination category as the first product that have similar purchase intent, and so on. Targeting similar and best-selling products helps maintain a product's industry ranking, while targeting combination products can achieve cross-category traffic growth.

[0064] Among them, specific feature tags can also be associated with corresponding product selection rules. When a user selects one or more feature tags, the system can select specific candidate second products based on the corresponding selection rules.

[0065] With the above-mentioned multiple feature tags provided, the specific backend interface can display, for example, Figure 3 The options shown at point 33 include "Similar Products", "Follow-up Investing in Popular Products" (corresponding to the feature tag of high-selling products), and "Bundled Products". Users can select from these options to determine their target for follow-up investment.

[0066] In practice, when selecting candidate second products based on the fuzzy configured follow-up target information, the number of products matching the selection rules may be large. To narrow down the range of candidate second products and achieve more accurate follow-up, the first user can be provided with operation options for configuring delivery preferences. This allows for the narrowing down of the range of candidate second products based on the corresponding delivery preference information. For example, see [link to relevant documentation]. Figure 3The "Set Follow-up Investment Preferences" option, shown at point 34, allows the first user to set their follow-up investment preferences. For example, clicking this option will take you to a preference settings interface or panel. In one specific implementation, this interface can look like this: Figure 4 As shown, users can configure corresponding follow-up preferences based on different characteristics. For example, for similar products, follow-up preferences can be configured from the perspective of price, whether to include products from this store, etc. For follow-up on popular products, follow-up preferences can be configured from the perspective of popular product type (whether it is a category popular product or a newly popular product in the category, etc.) and whether to include products from this store, etc. For bundled products, specific bundled categories can be configured (if the first user does not configure follow-up preferences, the system will determine the category that can be bundled with the specific first product based on specific algorithms, etc.).

[0067] In the aforementioned case of fuzzy follow-up targeting, since the first user's target is not specific to a particular product or brand, and the product database often contains a vast number of products, upon receiving the user's target information, an algorithmic model can be used to determine the range of candidate second data objects. Specifically, to improve recognition efficiency and accuracy, AI (Artificial Intelligence) models can be employed. These AI models typically refer to deep learning models containing massive amounts of parameters. Due to their large scale, these large AI models can store and process vast amounts of information, thereby achieving higher performance across various tasks.

[0068] In cases of similar product follow-up advertising, AI models can be used for similar product identification. For example, a prompt can be constructed based on the product information of the first product and the information of multiple products in the product database, and then input into the AI ​​model. The AI ​​model will then output similar products to the first product. If the first user has also configured advertising preferences, the products in the product database can be filtered based on these preferences before the AI ​​model identifies similar products from the remaining products. Alternatively, in cases of following up on best-selling products, products belonging to the same leaf category or similar range as the first product can be selected from the product database. Then, the historical sales data and other relevant information of these products can be input into the AI ​​model, which will then identify products that meet the criteria for best-selling products.

[0069] Furthermore, for cases involving bundled product pairings, AI models can also be used to determine which products can be paired with the first product, and from this, a range of candidate second products can be identified. Specifically, pairing relationships between different product categories can be predefined. Thus, when determining bundled products for the first product, target categories that can be paired with it can be identified based on the category to which the first product belongs. Products within these target categories are then filtered from the product information database. Finally, an AI model based on collaborative filtering algorithms is used to mine cross-category user association behavior data (e.g., if a user purchases product A and product B simultaneously, this constitutes association behavior data, etc.), and from this, a range of second products that can be paired with the first product can be identified, and so on.

[0070] Specifically, the AI ​​model used in this application embodiment can be a model that supports multimodal content processing capabilities (an AI technology that can understand or generate multiple forms of data (such as text, images, audio, etc.)). In specific implementation, some training data can be used to fine-tune the basic AI model to enhance the AI ​​model's capabilities in similar product recognition, popular product recognition, cross-category product recognition, etc.

[0071] In the case of precise follow-up investment, the first user can configure and specify the identifier of the product to be followed. That is, the follow-up target information can directly include the identifier of at least one specified product, and correspondingly, at least one specified product configured by the first user can be directly identified as a candidate second product. Specifically, when performing precise follow-up investment, the user can select... Figure 3 The "Custom Follow-up Products" option is shown at point 35. Then, by using the "Add Product" option, you can access the interface for specifying a particular product. For example, this interface can be like this: Figure 5 As shown, the first user can obtain relevant product information by searching in the search box shown in 51. This search supports multiple methods, including product title search, brand name search, store name search, and product ID search. After initiating a search, a list of products matching the search criteria will be displayed below the search box, and each product can have an "Add" option (e.g., ...). Figure 5 (As shown at point 52 in the image), this interface is used to set the corresponding product as the designated product. Additionally, the interface can display a list of added items, which users can edit, including removing previously added designated products, etc.

[0072] It's important to note that when the first user performs targeted follow-up advertising through customization, the specific products designated can be other products posted by the first user that have the same or similar sales attributes as the first product (usually so-called competing products). Alternatively, if the first product the first user wants to promote is a newly posted product (i.e., a new product), then products previously posted by the same first user, or related products previously posted by other first users, can also be designated as products to be followed up, and so on. In other words, the first user can choose to follow up on competing products to maintain their industry ranking, and can also help new products reach their target high-quality audience more quickly and accurately by having new products follow up on older products, and so on.

[0073] Another way to set follow-up targets for the first product is somewhere between fuzzy follow-up and precise follow-up. For example, you can set the brand name or store name information to be followed. In this case, the system can identify products associated with the brand name or store name that belong to the same leaf category as the first product as candidate second products.

[0074] Of course, in practical applications, other ways to set follow-up investment targets can also be supported, which will not be introduced one by one here.

[0075] S202: In the process of recalling and sorting data objects based on the data object search request of the second user or the need to recommend data objects to the second user, if at least one of the second data objects exists in the determined set of data objects to be displayed, then the target second data object is determined.

[0076] After identifying the range of candidate second products for the first product, follow-up product placement can be implemented in specific product search or recommendation scenarios. Specifically, when a second user initiates a product search request, or when product recommendations need to be made to a second user in a recommendation scenario, product recall can be performed, and the recalled products can be sorted. Different product recall strategies can be set for different scenarios. For example, in a search scenario, the product recall process can involve retrieving products related to the search keywords from the product database. This includes both advertised and organic products. Advertised products can include ordinary advertised products (including products where merchants manually purchase keywords for placement, or products where the system automatically purchases keywords for intelligent placement), and products intelligently placed using the follow-up placement method in this embodiment.

[0077] After the product recall is completed, the products can be sorted. Specifically, when sorting, for organic products, the product score is usually calculated based on the matching degree between the product and the keywords, the predicted conversion rate, etc. For ordinary advertised products, the base score can be adjusted by weighting the base score based on the merchant's bid information for the current keywords, etc., to obtain the final score. Alternatively, the system can dynamically bid based on the above base score to obtain the final score, and so on.

[0078] For products that are intelligently targeted using the follow-up bidding method in this application embodiment, although the bidding is also intelligently determined by the system, it differs from ordinary intelligently targeted advertising products in that, when intelligently bidding for a first product, it can first determine whether there is at least one of the aforementioned candidate second products in the set of products to be displayed. If there is, the target second product can be determined first, and the target second product can become the specific object to be followed. Then, the first product can be dynamically bid based on the situation of the target second product.

[0079] If the current set of products to be displayed includes only one of the aforementioned candidate second products, then that second product can be directly selected as the target second product. If the current set of products to be displayed includes multiple of the aforementioned candidate second products, then one of them can be selected as the target second product through random selection or other methods.

[0080] It should be noted that the specific sorting of recalled products may include a coarse sorting stage and a fine sorting stage. During the coarse sorting process, some products may be filtered out, and the fine sorting stage may further filter out some products. Therefore, in order to obtain more accurate information about the set of products to be displayed, it can be determined in the fine sorting stage whether there is at least one second product in the set of products to be displayed, and a dynamic bid can be made for the first product.

[0081] S203: Dynamically bid on the first data object based on the sorting position of the target second data object, so that the first data object is displayed near the resource position where the target second data object is located on the data object collection display page.

[0082] After identifying the target second product, a dynamic bid can be placed on the first product based on its sorting position. The goal of the bid is to display the first product near the resource slot where the target second product is located on the product collection display page.

[0083] In practice, the first product may appear directly in the set of products to be displayed, or it may not appear at all. In the latter case, since the second product appears in the set of products to be displayed, and the first product is configured to follow the bid for the second product, the first product can be added to the set of products to be displayed, and then a bid can be placed on the first product. That is, in this embodiment, even if the first product is not recalled by the recall algorithm (perhaps because the first product did not purchase a certain keyword / audience, etc.), as long as the second product it chose to follow is recalled, the first product has a chance to appear in the final set of products to be displayed. Furthermore, on the product set display page (including product search results pages or product recommendation pages, etc.), the first product can appear near the resource position where the second product is located. Here, "nearby" can include resource positions adjacent to the resource position where the target second product is located, or it can be several resource positions apart, as long as the probability of the first product being displayed on the same screen as the target second product is higher than a certain threshold. In this way, when a second user browses the target second product, they may also browse the first product. Since the first product and the target second product have some similar or identical sales attributes, this simultaneous display can make it more likely that the second user who is interested in the target second product will also click on the first product, and may even lead to a browse-to-purchase conversion. In other words, the target second product brings traffic to the first product.

[0084] Specifically, when dynamically bidding on the first product, the ranking position of the target second product can be determined first. Then, a trial bid can be placed on the first product, for example, starting with a relatively low price. The possible positions of the first product at the current price can be calculated. If it is far from the second product, the bidding strategy can be adjusted, including increasing the bid price, etc., until the first product can appear in a position near the second product. Of course, other dynamic bidding methods can also be used in practice, which will not be detailed here.

[0085] In summary, this application provides a "follow-up" intelligent ad placement method for merchants and other first users. In this method, the first user can configure follow-up target information for a first data object, and the system can determine the range of candidate second data objects based on the follow-up target information configured by the first user. Then, during the process of recalling and sorting data objects based on the second user's data object search request or the need to recommend data objects to the second user, if at least one second data object exists in the determined set of data objects to be displayed, the target second data object can be identified first. Then, the first data object is dynamically bid on based on the sorting position of the target second data object, so that the first data object is displayed near the resource position of the target second data object on the data object set display page. In this way, for data objects where it is difficult to determine a clear ad placement strategy or achieve good ad placement results from the data object's own perspective, following the second data object can be chosen, allowing the first data object to have the opportunity to gain exposure and conversion through the target second data object, thereby improving ad placement results. Of course, even if the data object itself does not have problems such as difficulty in determining an ad placement strategy, the solution provided in this application embodiment can still be used for follow-up ad placement.

[0086] It should be noted that the embodiments of this application may involve the use of user data. In practical applications, user-specific personal data may be used in the scheme described herein within the scope permitted by applicable laws and regulations, provided that it complies with the applicable laws and regulations of the country (e.g., with the user's explicit consent, with the user being properly notified, etc.).

[0087] Corresponding to the foregoing method embodiments, this application also provides a data object delivery device, which may include:

[0088] The candidate range determination unit is used to determine the range of candidate second data objects for the first data object based on the follow-up target information configured by the first user for the first data object;

[0089] The second data object determination unit is used to determine the target second data object in the process of recalling and sorting data objects based on the data object search request of the second user or the need to recommend data objects to the second user.

[0090] The dynamic bidding unit is used to dynamically bid on the first data object based on the sorting position of the target second data object, so that the first data object can be displayed near the resource position of the target second data object in the data object collection display page.

[0091] The target information for follow-up investment includes: data object feature information set according to the category to which the first data object belongs;

[0092] The candidate range determination unit can specifically be used for:

[0093] Based on the data objects in the data object information database that match the feature information, the range of candidate second data objects is selected.

[0094] Additionally, the device may also include:

[0095] The feature label providing unit is used to provide a variety of optional feature labels, so that the first user can express the feature information of the data object and determine the follow-up investment target information by selecting feature labels; wherein, the feature labels are defined in advance after classifying the data object follow-up investment targets in the follow-up investment scenario; the feature labels are associated with corresponding data object selection rule information.

[0096] The feature tags include: similar data objects, which are used to set the follow-up target information as follows: follow data objects under the leaf category to which the first data object belongs, whose similarity to the first data object meets the conditions, and whose prices are similar.

[0097] Alternatively, the feature tag includes: high-sales data objects, used to set the follow-up target information to: follow data objects under the leaf category to which the first data object belongs, whose sales volume and / or sales growth rate meet the conditions for deployment.

[0098] Alternatively, the feature tag includes: a matching data object, used to set the follow-up target information to: follow the matching category of the first data object for targeting data objects with similar purchasing intentions.

[0099] When using category-based matching, the candidate range determination unit can specifically be used for:

[0100] An AI model based on collaborative filtering algorithm is used to mine information from user association behavior data across categories to identify data objects with similar purchase intentions under the matching categories of the first data object, and these data objects are selected as the candidate second data objects.

[0101] Additionally, the device may also include:

[0102] A preference configuration option providing unit is used to provide operation options for configuring delivery preferences, so as to narrow down the range of candidate second data objects based on the corresponding delivery preference information.

[0103] In a specific implementation, the target information for follow-up investment may further include: identification information of at least one specified data object configured by the first user;

[0104] At this point, the candidate range determination unit can specifically be used for:

[0105] At least one specified data object configured by the first user is identified as a candidate second data object.

[0106] The specified data object includes: other data objects published by the first user that have the same or similar sales attributes as the first data object.

[0107] If the first data object is a newly published data object by the first user, then the specified data object includes: data objects previously published by the same first user, or data objects previously published by other first users that have the same or similar sales attributes as the first data object.

[0108] Furthermore, the target information for follow-up investment includes: brand name or store name information configured by the first user;

[0109] At this point, the candidate range determination unit can specifically be used for:

[0110] Data objects associated with the brand name or store name and belonging to the same leaf category as the first data object are identified as candidate second data objects.

[0111] If the sorting includes a coarse sorting stage and a fine sorting stage, then in the fine sorting stage, it can be determined whether there is at least one second data object in the set of data objects to be displayed, and a dynamic bid can be made for the first data object.

[0112] If the first data object is not recalled by the recall algorithm during the data object recall process, and there is at least one second data object in the set of data objects to be displayed, then the first data object is added to the set of data objects to be displayed, and a dynamic bid is made for the first data object so that the first data object is displayed near the resource position where the target second data object is located.

[0113] In addition, embodiments of this application also provide a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the method described in any of the foregoing method embodiments.

[0114] And an electronic device, comprising:

[0115] One or more processors; and

[0116] A memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the steps of the method described in any of the foregoing method embodiments.

[0117] A computer program product includes a computer program / computer executable instructions that, when executed by a processor in an electronic device, implement the steps of the method described in the foregoing method embodiments.

[0118] in, Figure 6 An exemplary architecture of an electronic device is shown, which may include a processor 610, a video display adapter 611, a disk drive 612, an input / output interface 613, a network interface 614, and a memory 620. The processor 610, video display adapter 611, disk drive 612, input / output interface 613, network interface 614, and memory 620 can communicate with each other via a communication bus 630.

[0119] The processor 610 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solution provided in this application.

[0120] The memory 620 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 620 can store the operating system 621 for controlling the operation of the electronic device 600, and the basic input / output system (BIOS) for controlling the low-level operations of the electronic device 600. Additionally, it can store a web browser 623, a data storage management system 624, and a data object placement system 625, etc. The aforementioned data object placement system 625 can be the application program that specifically implements the aforementioned steps in this embodiment. In summary, when implementing the technical solution provided in this application through software or firmware, the relevant program code is stored in the memory 620 and is called and executed by the processor 610.

[0121] Input / output interface 613 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touch screens, microphones, various sensors, etc., and output devices may include displays, speakers, vibrators, indicator lights, etc.

[0122] Network interface 614 is used to connect a communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0123] Bus 630 includes a pathway for transmitting information between various components of the device, such as processor 610, video display adapter 611, disk drive 612, input / output interface 613, network interface 614, and memory 620.

[0124] It should be noted that although the above-described device only shows the processor 610, video display adapter 611, disk drive 612, input / output interface 613, network interface 614, memory 620, bus 630, etc., in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the solution of this application, and does not necessarily include all the components shown in the figures.

[0125] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application.

[0126] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on its differences from other embodiments. In particular, for system or system embodiments, since they are fundamentally similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The systems and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. Components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0127] The data object delivery method and electronic device provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and its core ideas. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for deploying data objects, characterized in that, include: Based on the follow-up investment target information configured by the first user for the first data object, determine the range of candidate second data objects for the first data object; In the process of recalling and sorting data objects based on the data object search request of the second user or the need to recommend data objects to the second user, if at least one second data object exists in the determined set of data objects to be displayed, then the target second data object is determined. The first data object is dynamically bid on based on its sorting position, so that it is displayed near the resource position of the target second data object on the data object collection display page.

2. The method according to claim 1, characterized in that, The target information for follow-up investment includes: data object feature information set according to the category to which the first data object belongs; Determining the range of candidate second data objects for the first data object includes: Based on the data objects in the data object information database that match the feature information, the range of candidate second data objects is selected.

3. The method according to claim 2, characterized in that, Also includes: Multiple optional feature tags are provided so that the first user can express the feature information of the data object and determine the follow-up investment target information by selecting feature tags; wherein, the feature tags are defined in advance after classifying the data object follow-up investment targets in the follow-up investment scenario; the feature tags are associated with corresponding data object selection rule information.

4. The method according to claim 3, characterized in that, The feature tags include: similar data objects, which are used to set the follow-up target information as follows: follow data objects under the leaf category to which the first data object belongs, which meet the similarity conditions with the first data object, and which have similar prices for distribution.

5. The method according to claim 3, characterized in that, The feature tags include: high-sales data objects, which are used to set the follow-up target information to: follow the data objects under the leaf category to which the first data object belongs, whose sales volume and / or sales growth rate meet the conditions for deployment.

6. The method according to claim 3, characterized in that, The feature tags include: matching data objects, used to set the follow-up target information as: following data objects with similar purchasing intentions under the matching category of the first data object for targeting.

7. The method according to claim 6, characterized in that, The step of determining the range of candidate second data objects for the first data object based on the follow-up target information configured by the first user for the first data object includes: An AI model based on collaborative filtering algorithm is used to mine information from user association behavior data across categories to identify data objects with similar purchase intentions under the matching categories of the first data object, and these data objects are selected as the candidate second data objects.

8. The method according to claim 2, characterized in that, Also includes: Operational options are provided for configuring delivery preferences, so as to narrow down the range of candidate second data objects based on the corresponding delivery preference information.

9. The method according to claim 1, characterized in that, The target information for follow-up investment includes: identification information of at least one specified data object configured by the first user; Determining the range of candidate second data objects for the first data object includes: At least one specified data object configured by the first user is identified as a candidate second data object.

10. The method according to claim 9, characterized in that, The specified data object includes: other data objects published by the first user that have the same or similar sales attributes as the first data object.

11. The method according to claim 9, characterized in that, If the first data object is a newly published data object by the first user, then the specified data object includes: data objects previously published by the same first user, or data objects previously published by other first users that have the same or similar sales attributes as the first data object.

12. The method according to claim 1, characterized in that, The target information for follow-up investment includes: brand name or store name information configured by the first user; Determining the range of candidate second data objects for the first data object includes: Data objects associated with the brand name or store name and belonging to the same leaf category as the first data object are identified as candidate second data objects.

13. The method according to any one of claims 1 to 12, characterized in that, If the sorting includes a coarse sorting stage and a fine sorting stage, then in the fine sorting stage, it is determined whether there is at least one second data object in the set of data objects to be displayed, and a dynamic bid is made for the first data object.

14. The method according to any one of claims 1 to 12, characterized in that, If the first data object is not recalled by the recall algorithm during the data object recall process, and there is at least one second data object in the set of data objects to be displayed, then the first data object is added to the set of data objects to be displayed, and a dynamic bid is made for the first data object so that the first data object is displayed near the resource position where the target second data object is located.

15. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program performs the steps of the method described in any one of claims 1 to 14.

16. An electronic device, characterized in that, include: One or more processors; as well as A memory associated with the one or more processors, the memory being used 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 claims 1 to 14.

17. A computer program product comprising a computer program / computer executable instructions, characterized in that, When the computer program / computer executable instructions are executed by a processor in an electronic device, they implement the steps of the method according to any one of claims 1 to 14.

Images