Advertisement creating method and device, and storage medium

Abstract

The application discloses an advertisement creation method and device and a storage medium. The method comprises the following steps: obtaining a target product corresponding to an advertisement to be created; determining feature information and line information corresponding to the target product based on a preset database; determining a candidate object set by performing text splitting processing on the feature information and performing classification processing on a search object corresponding to the line information; determining a target delivery object set based on a delivery index and brand analysis data corresponding to each candidate object in the candidate object set; determining a bid of each target delivery object based on a bid reference range and a bid reference value of each target delivery object in the target delivery object set; and creating an advertisement according to the target delivery object set, the bid corresponding to the target delivery object, and a preset budget. Thus, the target delivery object and the bid corresponding thereto are reasonably determined based on key data corresponding to the target product in the preset database, the requirement for the experience of a merchant is reduced, and the advertisement promotion effect can be effectively improved.

Description

Methods, devices and storage media for creating advertisements Technical Field

[0001] This application relates to the field of advertising creation and processing technology, and more specifically, to an advertising creation method, apparatus and storage medium. Background Technology

[0002] E-commerce platforms provide a platform for users and merchants to conduct online transactions. Users can enter keywords of the products they want to buy to find the products they need; merchants can create advertisements to promote their products by setting corresponding keywords and bidding, so that the products can appear at the top of the display page when users enter keywords.

[0003] However, when creating an ad, the selection of keywords and the determination of bids rely heavily on the merchant's experience, making the process difficult and unable to guarantee the effectiveness of the advertising. Summary of the Invention

[0004] In view of the above problems, the present invention proposes an advertisement creation method, apparatus and storage medium to improve the above technical problems.

[0005] In a first aspect, embodiments of this application provide an advertisement creation method, which includes: obtaining the target product corresponding to the advertisement to be created; determining the feature information and product line information corresponding to the target product based on a preset database; determining a candidate object set by performing text splitting on the feature information and classifying the search objects corresponding to the product line information; determining a target target object set based on the placement indicators and brand analysis data corresponding to each candidate object in the candidate object set; determining the bid for each target target object based on the bid reference range and bid reference value for each target target object in the target target object set; and creating an advertisement based on the target target object set, the bid corresponding to the target target object, and a preset budget.

[0006] Secondly, embodiments of this application also provide an advertisement creation apparatus, which includes: an acquisition module for acquiring a target product corresponding to the advertisement to be created; a first determination module for determining feature information and product line information corresponding to the target product based on a preset database; a second determination module for determining a candidate object set by performing text splitting processing on the feature information and classifying search objects corresponding to the product line information; a third determination module for determining a target target object set based on the placement indicators and brand analysis data corresponding to each candidate object in the candidate object set; a bidding determination module for determining the bidding price of each target target object based on the bidding reference range and bidding reference value of each target target object in the target target object set; and a creation module for creating an advertisement based on the target target object set, the bidding price corresponding to the target target object, and a preset budget.

[0007] Thirdly, embodiments of this application also provide a computer-readable storage medium storing program code, wherein the above-described advertisement creation method is executed when the program code is run by a processor.

[0008] The technical solution provided by this invention specifically includes: acquiring the target product corresponding to the advertisement to be created; determining the feature information and product line information corresponding to the target product based on a preset database; determining a candidate object set by performing text splitting on the feature information and classifying the search objects corresponding to the product line information; determining a target target object set based on the placement indicators and brand analysis data corresponding to each candidate object in the candidate object set; determining the bid for each target target object based on the bid reference range and bid reference value for each target target object in the target target object set; and creating an advertisement based on the target target object set, the bid corresponding to the target target object, and a preset budget. Therefore, by reasonably determining the target target object and setting the corresponding bid based on the key data corresponding to the target product in the preset database, the requirements for merchant experience are reduced, the operational difficulty is decreased, and the advertising promotion effect can be effectively improved. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments and drawings obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0010] Figure 1 shows a flowchart of an advertisement creation method provided in an embodiment of this application.

[0011] Figure 2 shows a schematic diagram of the structure of an advertisement creation device provided in an embodiment of this application.

[0012] Figure 3 shows a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0013] Figure 4 shows a schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of this application. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0015] An e-commerce platform is a platform that allows users and merchants to conduct online transactions. Users can enter keywords of the products they want to buy on the e-commerce platform. The back-end operating system of the e-commerce platform determines the products associated with the keywords based on the keywords entered by the user and displays the products on the e-commerce platform's display page.

[0016] E-commerce platforms display a large number of products on their product pages, and products displayed at the top of the page have a greater chance of being purchased by users. Therefore, merchants can promote their products by setting relevant keywords and bidding on them to ensure that their products appear prominently on the display page when users enter those keywords.

[0017] Therefore, the accuracy of keyword selection and the corresponding bidding settings are two important aspects that affect the effectiveness of advertising promotion.

[0018] In related technologies, merchants typically choose keywords and set corresponding bids independently, and can also use timed tasks to verify whether the keyword selection and bid settings are reasonable.

[0019] However, the selection of keywords and the determination of corresponding bids largely depend on the merchant's experience. This is difficult for less experienced merchants to operate and they are prone to making unreasonable keyword selections and corresponding bid settings, which cannot guarantee the effectiveness of advertising.

[0020] To address the aforementioned issues, the inventors have proposed an advertisement creation method, apparatus, and storage medium as provided in this application. The method specifically includes: acquiring the target product corresponding to the advertisement to be created; determining the characteristic information and product line information corresponding to the target product based on a preset database; determining a candidate object set by performing text splitting on the characteristic information and classifying the search objects corresponding to the product line information; determining a target target object set based on the placement metrics and brand analysis data corresponding to each candidate object in the candidate object set; determining the bid for each target target object based on the bid reference range and bid reference value for each target target object in the target target object set; and creating the advertisement based on the target target object set, the bid corresponding to the target target object, and a preset budget.

[0021] Therefore, by rationally determining the target audience and setting corresponding bids based on key data corresponding to the target product in a pre-set database, the requirements for merchant experience are reduced, the operational difficulty is decreased, and the advertising promotion effect can be effectively improved. Please refer to the following steps for specific implementation details.

[0022] Please refer to Figure 1, which shows a flowchart of an advertisement creation method provided in an embodiment of this application. The method may include steps 110 to 160.

[0023] In step 110, the target product corresponding to the advertisement to be created is obtained.

[0024] Merchants can create advertising campaigns on e-commerce platforms. In the embodiments of this application, the advertising campaign to be created is an example of an SP (Sponsored Products) advertisement, and the e-commerce platform is an example of the Amazon platform.

[0025] The target product is the product associated with the advertising campaign to be created, used to determine the key information needed to create the advertising campaign.

[0026] In some implementations, step 110 may include the following steps.

[0027] (1) Obtain the product to be delivered corresponding to the advertisement to be created.

[0028] (2) Identify competing products that correspond to the product to be launched.

[0029] (3) Determine the target product based on the product to be launched and competing products.

[0030] The target product may include products that need to be promoted, or it may include competing products that are associated with the products to be promoted.

[0031] In some implementations, when creating an advertisement, the user selects the product to be promoted that corresponds to the advertisement to be created.

[0032] In some implementations, competing products can be determined based on the product to be launched; for example, competing products similar to the product to be launched can be determined based on the attributes, characteristics, and target consumer groups of the product to be launched.

[0033] In some implementations, users can also select competing products associated with the product to be advertised when creating an ad. For example, they can enter the ASIN (Amazon Standard Identification Number) code of the competing product. Users can determine competing products based on market research and data analysis of the product to be advertised. Competing products can be products belonging to the same category as the product to be advertised, have similar functions, have similar appearances, or have the same or similar target consumer groups as the product to be advertised. It is understood that this application is not limited to these.

[0034] When users do not conduct in-depth or comprehensive research or understanding of the products to be advertised, there may be discrepancies between the competing products identified by the users and the actual competing products of the product to be advertised. This can lead to significant errors in subsequent steps when determining candidate products based on the characteristics of the competing products.

[0035] To improve the accuracy of screening competing products, in some implementations, the step of determining competing products corresponding to the product to be launched may include the following steps.

[0036] (a) Obtain information about the product category to which the product to be launched belongs.

[0037] (b) In the preset database, query products that correspond to the category information and whose sales ranking data are within the preset ranking threshold as competing products.

[0038] In the embodiments of this application, the category information is the product category information. Optionally, the category information includes broad category information and sub-category information. The products are refined from broad category information to sub-category information. It can be understood that the more detailed the category information, the smaller the amount of product data associated with that category information, and the closer the similarity or association between the products associated with that category information.

[0039] Preferably, the category information is the subcategory information related to the product to be launched.

[0040] In some implementations, the preset database may include product-related information, such as category information and ranking information. The product information in the preset database can be obtained through the application programming interface (API) or report data provided by the e-commerce platform, or it can be obtained from the user's account data using the e-commerce platform account information provided by the user.

[0041] In some implementations, each product has a unique code, which allows you to retrieve information about that product from a pre-defined database. For example, you can use the product's code to search for the corresponding product category information in the pre-defined database.

[0042] In some implementations, the ranking information may include product sales ranking data. Products with higher sales volume are more competitive, so competing products can be selected based on sales ranking data.

[0043] Optionally, products belonging to the same category as the product to be advertised and whose sales ranking data falls within a preset ranking threshold can be selected as competing products. The preset ranking threshold can be set according to actual usage needs. For example, the preset ranking threshold can be set to 10%–30%, meaning products whose sales ranking data falls within the top 10%–30% can be selected as competing products.

[0044] This allows users to accurately and objectively identify competing products related to the product they are launching.

[0045] To further improve the accuracy of screening competing products, in some implementations, the competing products can be the result of deduplicating competing products selected by the user and those determined based on the category information corresponding to the product to be advertised. The specific descriptions of the user-selected products and the products determined based on the category information can be found in the foregoing embodiments.

[0046] It is understood that this application is not limited to this, and the determination of competing products can be adjusted according to the actual needs of the advertisement to be created. For example, competing products can be screened based on the products of the merchant's competitors, and this application does not restrict this.

[0047] In step 120, based on a preset database, the characteristic information and product line information corresponding to the target product are determined.

[0048] In the embodiments of this application, the preset database also stores feature information and product line information corresponding to the target product. The feature information and product line information corresponding to the target product can be queried in the preset database using the code corresponding to the target product.

[0049] Among them, feature information is information associated with the target product. Users can learn about the specific functions and features of the target product through feature information.

[0050] In some implementations, the feature information includes title information, which can be a description of the target product. The specific content of the title information can be set according to the attributes of the target product and to cater to user psychology. For example, the title information can be a combination of one or more pieces of information such as the efficacy information, trademark information, scope of application (guiding users to make correct purchases), or product characteristics (material, size, etc.). Taking the advertisement to be created as an SP advertisement as an example, the title information is the title of the target product corresponding to the advertisement to be created.

[0051] For example, the target product is a work boot, and the product title information can be "No Break In Period Soft Leather Work Boot" or "Soft Leather Work Boot with No Break-in Period", etc.

[0052] In some implementations, the title information may include key elements. After determining the target product for the advertisement, the merchant can describe the target product or set key elements to cater to user habits. For example, the merchant may choose a garment as the target product for the advertisement and determine "shirt" and "print" as key elements for this garment. These two key elements can be integrated into the title information of this garment, which could be "printed shirt" or "printed white shirt," etc. Taking the advertisement to be created as an SP advertisement as an example, the key elements are the keywords set for the advertisement campaign.

[0053] At the same time, merchants can also allocate corresponding advertising values ​​to key targets based on budget and the importance of key targets, so that when the relevant targets entered by users on the e-commerce platform match the key targets, the target products can appear in a prominent position on the e-commerce platform's display page.

[0054] In some implementations, the feature information includes selling point information, which can be a characteristic of the target product. The specific content of the selling point information can be set according to the features of the target product and to cater to user habits. Similarly, selling point information can also be a combination of one or more information such as the target product's efficacy information, trademark information, scope of application (guiding users to make correct purchases), or product characteristics (material, features, size, etc.). Taking the advertisement to be created as an SP advertisement as an example, the selling point information is the selling point of the target product corresponding to the advertisement to be created.

[0055] In some implementations, the selling point information may include key objects, and the specific description of the key objects can be referred to in the foregoing embodiments.

[0056] The difference between title information and selling point information is that the main purpose of title information is to directly reflect the core features of the target product and attract users to the link where the target product is located to learn about it, while the main purpose of selling point information is to introduce the main features of the target product and attract users to buy it.

[0057] It is understood that this application is not limited to this, and the feature information can be adjusted accordingly based on the actual situation of the advertisement to be created. For example, the feature information includes descriptive information, which can be a specific description of the target product. Taking the advertisement to be created as an SP advertisement as an example, the descriptive information is a description of the target product corresponding to the advertisement to be created.

[0058] The product line information includes related products to the target product and the corresponding search objects for these related products. These related products can be determined by the merchant based on the target product. For example, if the target product is a white thermos cup, then the related products could be thermos cups from other brands, thermos cups of different sizes, thermos cups with different opening methods, and glass cups, etc.

[0059] The search term is randomly entered by the user based on their needs when searching for products on an e-commerce platform. For example, if a user enters "shirt" into the search query on an e-commerce platform, and the platform's display page shows the corresponding products from the product line information, then "shirt" is the search term for that product. Taking a SP (Special Offer) advertising campaign as an example, the search term is the search keyword corresponding to the product in the product line information.

[0060] Because the amount of data on search targets for products in the product line information can be quite large, the computational workload for determining candidate products from these search targets in subsequent steps is substantial. Furthermore, due to factors such as product improvements, the reference value of some search targets may gradually decrease over time. Therefore, the preferred search targets are those corresponding to products in the product line information within a preset time period. For example, the search targets might be those corresponding to products in the product line information within the past 30 days.

[0061] In step 130, a candidate object set is determined by performing text splitting on the feature information and classifying the search objects corresponding to the product line information.

[0062] When users search for products on e-commerce platforms, they typically enter a search term. The characteristic information of the target product is designed to help users quickly understand its general information. The search term is highly likely to appear in the characteristic information of the target product. However, since the characteristic information can be quite extensive, if merchants were to advertise every single piece of information in the characteristic information to increase the likelihood of the target product appearing on the e-commerce platform's display page and in a prominent position, the cost would be high. Furthermore, due to the lack of targeting, there might even be a negative correlation between advertising expenditure and advertising revenue. In reality, only a portion of the characteristic information of the target product will match the search term. Therefore, selecting the information in the characteristic information of the target product that matches the search term and targeting it as the key information for advertising can effectively improve the advertising promotion effect of the target product.

[0063] In the embodiments of this application, at least one key object is selected from the feature information by performing text splitting processing on the feature information, and these key objects are used as candidate objects for subsequent ad delivery of the ad to be created.

[0064] As can be seen from the aforementioned implementation methods, the products in the product line information are products related to the target product. The search objects corresponding to these products have been verified by user practice. Therefore, selecting search objects with high user usage from these search objects has high reference value for setting key objects. These search objects with high user usage can be used as candidate objects for subsequent advertising placement of the advertisement to be created.

[0065] The candidate object set is the collection of the aforementioned candidate objects, that is, the collection of candidate objects obtained from the feature information and the candidate objects selected from the search objects corresponding to the products in the product line information.

[0066] In some implementations, step 130 may include the following steps.

[0067] (1) Perform text segmentation on the feature information and classify the split objects into a candidate object set.

[0068] In embodiments of this application, text segmentation processing includes splitting feature information into individual objects and combining the individual objects into objects of different word lengths as candidate objects.

[0069] For example, if the feature information is "No Break In Period Soft Leather Work Boot", it can be broken down into individual objects "No", "Break", "In", "Period", "Soft", "Leather", "Work", and "Boot". These individual objects can then be combined into objects of different word lengths. For instance, combining the individual objects "No", "Break", "In", "Period", "Soft", "Leather", "Work", and "Boot" into candidate objects with a word length of 2 results in the candidate objects "No Break", "Break In", "In Period", "Period Soft", "Soft Leather", "Leather Work", and "Work Boot". These candidate objects "No Break", "Break In", "In Period", "Period Soft", "Soft Leather", "Leather Work", and "Work Boot" can then be added to the candidate object set.

[0070] In some implementations, the title information is split into individual objects, and the individual objects are combined into objects of different word lengths as candidate objects. For example, if the title is "No Break In Period Soft Leather WorkBoot", it can be broken down into individual objects: "No", "Break", "In", "Period", "Soft", "Leather", "Work", and "Boot". These individual objects can then be combined into objects of different word lengths. For instance, combining the individual objects "No", "Break", "In", "Period", "Soft", "Leather", "Work", and "Boot" into candidate objects with a word length of 3 results in the candidate objects "No Break In", "Break In Period", "InPeriod Soft", "Period Soft Leather", "Soft Leather Work", and "Leather Work Boot". These candidate objects "No Break In", "Break In Period", "In Period Soft", "Period SoftLeather", "Soft Leather Work", and "Leather Work Boot" can then be added to the candidate object set.

[0071] In some implementations, the selling point information is split into multiple objects with different preset word lengths; all the split objects are then grouped into a candidate object set. For example, if the selling point is "No Break In Period Soft Leather Work Boot", it can be broken down into individual objects: "No", "Break", "In", "Period", "Soft", "Leather", "Work", and "Boot". These individual objects can then be combined into objects of different word lengths. For instance, combining the individual objects "No", "Break", "In", "Period", "Soft", "Leather", "Work", and "Boot" into candidate objects with a word length of 5 results in the candidate objects "No Break In Period Soft", "BreakIn Period Soft Leather", "In Period Soft Leather Work", and "Period Soft Leather Work Boot". These candidate objects "NoBreak In Period Soft", "Break In Period Soft Leather", "In Period Soft Leather Work", and "Period Soft Leather Work Boot" can then be added to the candidate object set.

[0072] In some implementations, text segmentation can also utilize part-of-speech analysis and semantic analysis to pre-process feature information. For example, by removing prepositions, articles, and interjections from the feature information, preliminary feature information can be obtained. This preliminary feature information can then be split into multiple individual objects, and these individual objects can be combined into objects of different word lengths as candidate objects. This can make the candidate objects obtained after text segmentation more readable or reliable.

[0073] (2) Classify the search objects corresponding to the product line information, and assign the search objects belonging to the target category to the candidate object set.

[0074] Since the data volume of the search objects corresponding to product line information may be quite large, or some search objects may perform poorly (low user search volume) and not be of reference value, the search objects are classified to select those with high user usage or good performance as candidate objects.

[0075] Furthermore, the step involves categorizing the search objects corresponding to the product line information, identifying the search objects belonging to the target category, and thus determining the candidate objects. This may specifically include the following steps.

[0076] (a) Based on the usage benefit index of the search objects corresponding to the product line information, the search objects are classified.

[0077] In embodiments of this application, the effectiveness metrics may include advertising expenditure ratio data and order volume data. The advertising expenditure ratio data indicates the ratio between advertising spending and advertising sales revenue for the product corresponding to the search target, while the order volume data indicates the sales volume of the search target within a preset time period. Both the advertising expenditure ratio data and the order volume data can reflect the user usage rate and performance of the search target.

[0078] In some implementations, the effectiveness indicators may also include rankings from brand analytics data. Brand analytics data consists of search terms entered by users through an e-commerce platform, the number of searches for those terms, and the ranking of the search terms based on the search volume. This brand analytics data is stored in a pre-defined database, and its ranking can be obtained by inputting a search term into the database. Taking Amazon as an example, the brand analytics data is Brand Analytics (ABA).

[0079] It is understood that this application is not limited to this, and the effectiveness indicators may also include indicators such as the exposure data of the products corresponding to the search target, which can be adjusted according to actual needs.

[0080] In embodiments of this application, the categories of search objects may include high-traffic object categories, no-conversion object categories, low-conversion object categories, high-conversion object categories, and potential object categories. More specifically, this is further explained below with reference to Table 1 - Search Object Category Allocation Table.

[0081] Search category ranking, order volume data, preset advertising spending percentage. High-traffic categories with no conversions (top 50); low-conversion categories (greater than 0, greater than 50%); potential categories (greater than 0 and less than 1, less than or equal to 50%); high-conversion categories (greater than or equal to 1, less than or equal to 50%). surface

[0082] Table 1 - Category Assignment Table for Search Targets

[0083] The "high-traffic object" category is a collection of search objects whose corresponding ranking in brand analysis data is higher than a preset ranking. For example, if the preset ranking is the top 50, then search objects related to product line information that rank in the top 50 in brand analysis data will be classified as high-traffic objects.

[0084] The "no conversion" category is the set of search objects whose corresponding order volume data equals the second preset order volume. For example, if the second preset order volume is 0, then search objects with an order volume data of 0 are classified as "no conversion" objects.

[0085] The low-conversion category is a set of search objects whose corresponding advertising spending percentage is greater than a preset advertising spending percentage, and whose corresponding order volume is greater than a second preset order volume. For example, if the preset advertising spending percentage is 50% and the second preset order volume is 0, then search objects whose corresponding advertising spending percentage is greater than 50% and whose corresponding order volume is greater than 0 are classified as low-conversion objects.

[0086] The potential object class is a set of search objects whose corresponding advertising spending percentage is less than or equal to the preset advertising spending percentage, and whose corresponding order volume is greater than the second preset order volume and less than the first preset order volume. For example, if the preset advertising spending percentage is 50%, the second preset order volume is 0, and the first preset order volume is 1, then search objects whose corresponding advertising spending percentage is less than or equal to 50% and whose corresponding order volume is greater than 0 and less than 1 are classified as potential objects.

[0087] The high-conversion target category is a set of search objects whose corresponding advertising spending percentage is less than or equal to a preset advertising spending percentage, and whose corresponding order volume is greater than or equal to a first preset order volume. For example, if the preset advertising spending percentage is 50% and the first preset order volume is 1, then search objects whose corresponding advertising spending percentage is less than or equal to 50% and whose corresponding order volume is greater than or equal to 1 are classified as high-conversion targets.

[0088] (b) Assign search objects belonging to the target category to the candidate object set.

[0089] In the embodiments of this application, the target categories are high-traffic object category, high-conversion object category and potential object category. After classifying the search objects corresponding to the products in the product line information, the search objects belonging to the target category are taken as candidate objects and added to the candidate object set, so as to filter out the search objects corresponding to the product line information with reference value and add them to the candidate object set, thereby reducing the amount of calculation in subsequent steps.

[0090] In summary, the candidate object set is a collection of candidate objects belonging to the target category (high-traffic object category, high-conversion object category, and potential object category) after the multiple candidate objects obtained by text segmentation of feature information and the search objects corresponding to products in product line information are classified.

[0091] In step 140, the target target audience set is determined based on the advertising metrics and brand analysis data corresponding to each candidate in the candidate audience set.

[0092] In the embodiments of this application, the targeting metrics may include the word frequency, word length, and brand information of the target product for each candidate object. The word frequency represents the frequency of a candidate object's appearance in the candidate object set. Since the aforementioned implementation methods show that the feature information may contain search objects corresponding to products in the product line information, the candidate objects obtained after text segmentation processing of the feature information may be the same as search objects belonging to the target category corresponding to products in the product line information. Therefore, the same candidate object may appear multiple times in the candidate object set. The word length represents the length of the candidate object. For example, if the candidate object is "led light strip," then the word length of the candidate object is 3; if the candidate object is "printed shirt," then the word length of the candidate object is 4. The brand information indicates the company name or product name of the target product, etc. The brand information can be queried in a preset database using the coding information corresponding to the target product.

[0093] As can be seen from the aforementioned implementation method, the brand analysis data consists of the search object entered by the user through the e-commerce platform, the number of searches corresponding to the search object, and the ranking of the search object after sorting the search objects according to the number of searches. The brand analysis data is stored in a preset database, and the ranking of the search object can be obtained by entering the search object into the preset database.

[0094] In some implementations, step 140 may include the following steps.

[0095] (1) Based on the brand information of the target product, brand analysis data, word length, word frequency and text information of the candidate objects, the candidate objects are classified and the target object set is determined.

[0096] Candidates are categorized based on brand information, brand analysis data, keyword length, keyword frequency, and textual information of the target product. Candidate categories can include long-tail objects, attribute objects, brand objects, traffic objects, and other object categories. More specifically, this is further elaborated in Table 2 – Candidate Category Allocation Table.

[0097]

[0098] Table 2 - Category Allocation Table of Candidates

[0099] The long-tail object class consists of a set of candidate objects whose corresponding word frequency is greater than the target number, whose corresponding word length is greater than or equal to the first preset word length, and which do not exist in the brand analysis data; and a set of candidate objects whose corresponding word length is greater than or equal to the first preset word length and which exist in the brand analysis data. The target number is the total number of target products. A candidate object whose corresponding word frequency is greater than or equal to the target number can be considered to have appeared at least once in the product search objects in the feature information (title information or selling point information) or product line information corresponding to each target product. A candidate object that meets this condition can be considered a general search object.

[0100] The attribute object class is a collection of candidate objects whose corresponding text information contains preset attribute words, whose word length is less than or equal to a second preset word length, and which exist in the brand analysis data. The second preset word length is less than the first preset word length; for example, the second preset word length is 3 and the first preset word length is 4. The preset attribute words can be words that reflect the attributes of the candidate objects. For example, the preset attribute words can be words such as "APP", "RGB", or "number + ft".

[0101] The brand object class is a collection of candidate objects whose corresponding text information contains brand information.

[0102] The Traffic Object class is a collection of candidate objects that rank high in the brand analytics data according to a preset traffic ranking. For example, if the preset traffic ranking is 1000, then the candidate objects ranked within the top 1000 in the brand analytics data can be classified as Traffic Objects.

[0103] Other object classes are the collection of candidate objects whose corresponding word frequency is less than the target number and do not exist in the brand analysis data; and the collection of candidate objects whose corresponding text information does not contain preset attribute words, whose corresponding word length is less than or equal to the first preset word length and exist in the brand analysis data.

[0104] Candidates belonging to the target categories are grouped into the target set. The target categories are high-traffic target, high-conversion target, and potential target. Candidates belonging to the target categories (high-traffic target, high-conversion target, and potential target) are selected as targets and added to the target set. Candidates with reference value or high reference value are selected and added to the target set.

[0105] (2) Based on brand analysis data and the word frequency of candidate words, sort the target audience in the target audience set to determine the target audience set.

[0106] In the embodiments of this application, candidate objects belonging to the targeting categories (high-traffic target objects, high-conversion target objects, and potential target objects) are sorted according to their ranking in brand analysis data. When different candidate objects have the same ranking, they are sorted according to the frequency of their corresponding keywords to determine the target target object set. It is worth noting that since the target target object set may have a large number of objects, merchants can select the corresponding number of objects from the target target object set in order for advertising based on actual needs and other factors.

[0107] In step 150, the bid for each target target is determined based on the bid reference range and bid reference value for each target target in the target target set.

[0108] When merchants select a corresponding number of target audiences from a set of target audiences for advertising based on actual needs and other factors, they need to match corresponding bids for the selected target audiences. The bid for each target audience ensures that when a user's search query on the e-commerce platform matches the target audience, the corresponding target product will appear prominently on the e-commerce platform's display page. Taking an ad to be created as an SP ad as an example, the bid is the keyword bid (Bid) for the target product corresponding to the ad.

[0109] It is worth noting that the bid for the target product and the corresponding target audience does not necessarily mean that the higher the bid, the better the ad performance of the target audience. Therefore, it is crucial to match an appropriate bid for the target audience.

[0110] In embodiments of this application, matching an appropriate bid to each target audience may include the following steps.

[0111] (1) Determine the bidding reference range for the target audience based on whether there is related historical advertising data for the products in the product line information and the historical advertising data of the online stores associated with the products to be advertised.

[0112] When determining the bid for a target product associated with a target audience, historical advertising data of the target product can be used. Since historical advertising data has been validated by user experience, determining the bid based on historical advertising data has relatively high reference value. However, when the target product is a new product, there is no historical advertising data available. In this case, it is impossible to determine the bid based on the target product's historical advertising data. To address this issue, this application provides the following three solutions.

[0113] (a) If there is associated historical advertising data for the products in the product line information, the first preset range of average cost-per-click for the corresponding products with associated historical advertising data in the product line information shall be used as the bidding reference range for the target audience.

[0114] Based on the historical advertising data of products (products related to the target product) in the product line information, a reference range for bidding on the corresponding target audience is determined. Since the products in the product line information are extremely similar to the target product in terms of product attributes, functions, or target customer groups, the historical advertising data of the products in the product line information has some reference value for determining the bidding value of the target audience. More specifically, the reference range for bidding is determined based on the historical advertising data of products that have advertised in the product line information and a first preset range.

[0115] Historical advertising data can include total advertising spending and ad clicks. Total advertising spending indicates the total advertising expenditure of products that have run ads in the product line information associated with the target product. Ad clicks indicate the total number of clicks for products that have run ads in the product line information associated with the target product. Based on the total advertising spending and total clicks, the average advertising spending and average clicks are determined respectively. Then, based on the ratio of the average advertising spending to the average clicks, the average cost per click (CPC) is determined.

[0116] Determine the bidding reference range based on the average cost-per-click (CPC) and the first preset range. For example, if the average CPC is 1 and the first preset range is [50%, 80%], then the bidding reference range is [0.5, 0.8]. Understandably, the larger the first preset range, the larger the bidding reference range. Choosing an appropriate bidding reference range can provide a good reference for determining the bidding value for the target audience in subsequent steps.

[0117] In some implementations, businesses can determine the specific value of the first preset range based on their business experience.

[0118] In some implementations, a neural network model can be established, and the historical advertising data corresponding to the products in the product line information can be used as a training dataset. The neural network model can be trained based on the training dataset so that the neural network model can predict the optimal first preset range of values.

[0119] (b) If there is no associated historical advertising data for the product in the product line information, the second preset range of average cost-per-click for the corresponding products in the associated online stores of the product to be advertised shall be used as the bidding reference range for the target advertising object.

[0120] As can be seen from the aforementioned implementation method, the bidding reference range of the target target can be determined based on the historical advertising data of the corresponding product in the product line information. However, when none of the products related to the target product have been advertised, that is, when there is no historical advertising data for the products related to the target product, it is impossible to determine the bidding reference range of the target target by using the historical advertising data associated with the product in the product line information.

[0121] Generally, an online store on an e-commerce platform will sell similar products (such as different styles of clothing), products belonging to the same brand, or products belonging to the same category (such as home appliances). In other words, there is a certain correlation between different products in the same online store. Therefore, the products that have been advertised in the online store and their related historical advertising data have certain reference value for determining the bidding reference range of the target audience.

[0122] More specifically, the bidding reference range is determined based on historical advertising data of products already advertised in the target product's (the product to be advertised) associated online stores and a second preset range. Historical advertising data can include total advertising expenditure and ad clicks. Total advertising expenditure indicates the total advertising expenditure of products already advertised in the target product's associated online stores, and ad clicks indicate the total number of clicks on products already advertised in the target product's associated online stores. Based on the total advertising expenditure and total clicks, the average advertising expenditure and average clicks are determined respectively. Finally, the average cost-per-click (CPC) is determined based on the ratio of the average CPC to the average clicks.

[0123] Determine the bidding reference range based on the average cost-per-click (CPC) and the second preset range. For example, if the average CPC is 1 and the second preset range is [50%, 80%], then the bidding reference range is [0.5, 0.8]. Understandably, the larger the second preset range, the larger the bidding reference range. Choosing an appropriate bidding reference range can provide a better reference for determining the bidding for the target audience in subsequent steps.

[0124] In some implementations, businesses can determine the specific value of the second preset range based on their business experience.

[0125] In some implementations, a neural network model can be established, and the historical advertising data corresponding to the products in the product line information can be used as a training dataset. The neural network model can be trained based on the training dataset so that the neural network model can predict the optimal second preset range of values.

[0126] (c) If there is no associated historical advertising data for the products in the product line information and there is no historical advertising data for the products in the associated online stores of the products to be advertised, then the bidding reference range corresponding to the target audience shall be determined based on the pricing of the products to be advertised, the estimated percentage of advertising expenditure, and the estimated conversion rate.

[0127] As described in the aforementioned implementation methods, the bidding reference range for the target audience can be determined based on the historical advertising data corresponding to the products in the product line information, or based on the historical advertising data corresponding to products that have already been advertised in the online stores associated with the target product. When neither of these two conditions is met, the bidding reference range for the target audience can be determined based on the corresponding pricing of the product to be advertised, the estimated percentage of advertising expenditure, and the estimated conversion rate.

[0128] The estimated advertising cost of sales (ACOS) measures the cost of advertising a product; the lower the ACOS value, the lower the predicted advertising cost. The estimated conversion rate (Cr) predicts the probability that a user will click on a product link, view the product, and then choose to purchase it. By building a neural network model and training it with historical advertising data from products that have already been advertised, the model can predict the estimated advertising cost of sales and the estimated conversion rate for the product to be advertised, based on its attributes, pricing, and other information.

[0129] More specifically, the first calculation volume is determined by multiplying the estimated advertising expenditure percentage by the first preset percentage; the second calculation volume is determined by multiplying the estimated conversion rate by the second preset percentage; the average cost-per-click for the target audience is determined by multiplying the corresponding price of the product to be advertised, the first calculation volume, and the second calculation volume; and the bidding reference range for the target audience is determined by multiplying the average cost-per-click by the third preset threshold.

[0130] Merchants can determine the specific values ​​of the first and second preset percentages based on their business experience.

[0131] (2) Obtain the bidding reference value of the target target; determine the bidding price of the target target based on the bidding reference range and bidding reference value corresponding to the target target.

[0132] In the embodiments of this application, the bidding reference value is the theoretical bidding price matched by the system for each target object based on data analysis technology and neural network technology.

[0133] When the bid reference value corresponding to the target object is within the bid reference range, the bid of the target object is the bid reference value; when the bid reference value corresponding to the target object is not within the bid reference range, the bid of the target object is the boundary of the bid reference range.

[0134] For example, assuming the reference bid value for the target audience is 0.7 and the reference bid range is [0.5, 0.8], then the bid for the target audience is 0.7; assuming the reference bid value for the target audience is 0.4 and the reference bid range is [0.5, 0.8], then the bid for the target audience is 0.5; and assuming the reference bid value for the target audience is 1 and the reference bid range is [0.5, 0.8], then the bid for the target audience is 0.8.

[0135] It's worth noting that the bid for the target audience is dynamically adjusted based on the bidding activities of other merchants. For example, when a user enters a search term matching the target audience on an e-commerce platform, two products, Product A and Product B, appear on the platform's display page. Product A's target audience has a bid of 1 yuan, while Product B's target audience has a bid of 3 yuan. Product B will appear before Product A on the e-commerce platform's display page. However, although Product B is positioned earlier than Product A, its target audience has a relatively higher bid. Therefore, the bid for the target audience A associated with Product B can be appropriately reduced.

[0136] In step 160, an advertisement is created based on the target audience set, the bid corresponding to the target audience, and the preset budget.

[0137] In the embodiments of this application, the preset budget is the daily budget amount for advertising, and the preset budget can be set by the merchant.

[0138] The types of ads to be created can include automatic ads, exact match ads, and phrase match ads. Merchants can choose at least one of these ad types according to their actual needs.

[0139] In automatic advertising, the matching method between the target audience and the user's search query is broad match. More specifically, broad match breaks down the target audience into individual objects, and the search query includes all or some of the objects within the target audience. For example, if the target audience is "led light," it can be broken down into "led" and "light." When the search query is "led light," since it includes all objects within the target audience (i.e., both "led" and "light"), a user searching for "led light" on an e-commerce platform will find the target product using broad match. Similarly, if the search query is "led," since it includes some objects within the target audience, a user searching for "led" on an e-commerce platform will also find the target product using broad match. Understandably, search queries matched using broad match have a lower relevance to the target audience, potentially resulting in fewer user purchases.

[0140] Exact match advertising refers to a matching method where the target audience and the user's search query are exactly the same. More specifically, exact match means the search query and the target audience are identical. For example, if the target audience is "led light" and the search query is "led light," then when a user searches for products using "ledlight" on an e-commerce platform, the target audience "led light" and the search query "led light" will be matched using exact match, and the target product will appear on the e-commerce platform's display page. Conversely, if the search query is simply "led," because the search query "led" and the target audience "led light" are not exactly the same, the search query "led" will not find the target product using exact match.

[0141] Phrase match advertising refers to a matching method where the target audience and the user's search query are matched using phrase match. More specifically, phrase match means the search query contains complete information about the target audience. For example, if the target audience is "led light" and the search query is "led light strip," because the search query "ledlight strip" contains complete information about the target audience "led light," when a user searches for products using "led light strip" on an e-commerce platform, the target audience "led light" and the search query "ledlight strip" can be matched using phrase match, and the target product corresponding to "led light" will appear on the e-commerce platform's display page. Conversely, if the target audience is "led light" and the search query is "led strip light," when a user searches for products using "led strip light" on an e-commerce platform, because the search query "led strip light" does not contain complete information about the target audience "led light," the target audience "led light" and the search query "led strip light" cannot be matched using either phrase match or exact match, and the target product associated with the target audience will not appear on the e-commerce platform's display page.

[0142] To ensure a high relevance between the search object and the target audience, it is preferable to use an exact match mode or a phrase match mode when matching the search object and the target audience.

[0143] Furthermore, the matching method between the search object and the target object is determined based on the word length corresponding to the target object. More specifically, when the word length corresponding to the target object is less than the second preset word length, the preferred matching method for the target object is phrase matching; when the word length corresponding to the target object is greater than or equal to the second preset word length, the preferred matching method for the target object is exact matching.

[0144] The specific value of the second preset word length can be determined based on user habits. For example, the second preset word length can be 3. When the word length corresponding to the target audience is greater than or equal to 3, the matching method for the target audience is preferably exact match mode; when the word length corresponding to the target audience is less than 3, the matching method for the target audience is preferably phrase match mode.

[0145] In embodiments of this application, a preset budget can be allocated to different types of advertisements according to the types of advertisements selected by the merchant. For example, the preset budget can be allocated to the types of advertisements selected by the merchant according to an average allocation ratio.

[0146] Therefore, when merchants determine the target audience associated with the product to be advertised and set the corresponding bids for the target audience, the competing products of the product to be advertised provide a certain reference value for setting the target audience associated with the product to be advertised. Thus, the target product is determined based on the product to be advertised and its competitors. Furthermore, the target product's corresponding feature information and product line information are queried in a pre-set database. The feature information is then processed through text segmentation to obtain multiple candidate objects (which serve as candidates for the target audience associated with the product to be advertised). Finally, the search objects corresponding to the products in the product line information are further... The process involves initial screening of search objects corresponding to products in the product line information. Search objects belonging to the target categories (high-traffic, high-conversion, and potential categories) are categorized into candidate objects, thus determining the candidate object set. Further, candidate objects belonging to the advertising categories (high-traffic, high-conversion, and potential categories) within the candidate object set are assigned to the target object set for advertising. Based on advertising metrics and brand analysis data, the target advertising objects with higher reference value within the target advertising object set are determined. Then, based on the bidding reference range and value, the bid for each target advertising object within the target advertising object set is determined. Merchants can then select the corresponding number of target advertising objects from the target advertising object set as the target advertising objects associated with the products to be advertised, completing the ad creation.

[0147] Please refer to Figure 2, which illustrates an advertisement creation apparatus provided in an embodiment of this application. The advertisement creation apparatus 200 includes: an acquisition module 210, a first determination module 220, a second determination module 230, a third determination module 240, a bidding determination module 250, and a creation module 260. Specifically:

[0148] The acquisition module 210 is used to acquire the target product corresponding to the advertisement to be created.

[0149] In some embodiments, the acquisition module 210 may further include:

[0150] The "Products to be Deployed" module is used to obtain the products to be deployed corresponding to the advertisement to be created.

[0151] The Level 1 Competitive Product Determination Module is used to determine the competing products corresponding to the product to be launched.

[0152] In some embodiments, the competing product determination module may further include:

[0153] The category information determination module is used to obtain the category information of the product to be deployed;

[0154] The secondary competitive product determination module is used to query the preset database for products whose sales ranking data corresponding to the category information is within a preset ranking threshold as competitive products.

[0155] The target product determination module is used to determine the target product based on the product to be launched and the competing products.

[0156] The first determining module 220 is used to determine the feature information and product line information corresponding to the target product based on a preset database.

[0157] The feature information includes title information and selling point information.

[0158] In some embodiments, the first determining module 220 may further include:

[0159] The feature information determination module is used to determine the title information and selling point information corresponding to the target product based on a preset database.

[0160] The second determining module 230 is used to determine a candidate object set by performing text splitting processing on the feature information and classifying the search objects corresponding to the product line information.

[0161] In some embodiments, the second determining module 230 may further include:

[0162] The first candidate object set determination module is used to perform text segmentation processing on the feature information and assign the split objects to the candidate object set.

[0163] The first splitting module is used to split the title information and the selling point information into multiple objects with different preset word lengths;

[0164] The first candidate object set determination submodule is used to assign all the split objects to the candidate object set.

[0165] The second candidate object set determination module is used to classify the search objects corresponding to the product line information, and assign the search objects belonging to the target category obtained by classification to the candidate object set.

[0166] In some embodiments, the second candidate object set determination module may include:

[0167] The first classification module is used to classify the search objects based on the usage benefit indicators of the search objects corresponding to the product line information.

[0168] The second candidate object set determination submodule is used to classify search objects belonging to the target category into the candidate object set;

[0169] The target categories include at least one of high-conversion target categories, high-traffic target categories, and potential target categories; the usage efficiency indicators include: advertising expenditure ratio data, order volume data, and ranking in brand analysis data.

[0170] In some embodiments, the second candidate object set determination submodule may include:

[0171] The high-conversion object class determination module is used to determine that the search object belongs to the high-conversion object class if the advertising spending percentage corresponding to the search object is less than or equal to a preset advertising spending percentage and the corresponding order volume is greater than or equal to a first preset order volume; or,

[0172] The potential keyword category determination module is used to determine that the search object belongs to a potential keyword category if the advertising spending percentage corresponding to the search object is less than or equal to the preset advertising spending percentage, and the corresponding order volume is greater than a second preset order volume, and the corresponding order volume is less than a first preset order volume; or,

[0173] The high-traffic object class determination module is used to retrieve the ranking of the number of searches associated with the search object based on brand analysis data in a preset database; if the ranking of the search object is higher than the preset ranking, then the search object is determined to belong to the high-traffic object class.

[0174] The third determining module 240 is used to determine the target target object set based on the placement indicators and brand analysis data corresponding to each candidate object in the candidate object set.

[0175] The targeting metrics include the word frequency, word length, and brand information of the target product for each candidate; the word frequency represents the frequency with which the candidate appears in the candidate set.

[0176] The target audience determination module is used to classify the candidate objects based on the brand information of the target product, brand analysis data, word length, word frequency and text information corresponding to the candidate objects, and determine the target audience set.

[0177] In some embodiments, the module for determining the set of objects to be delivered may further include:

[0178] The candidate object classification module is used to classify the candidate objects based on the brand information of the target product, brand analysis data, word length, word frequency and text information of the candidate objects;

[0179] The "Target Object Set Determination" submodule is used to assign candidate objects belonging to the target object set to the target object set.

[0180] The target audience set determination module is used to sort the target audiences in the target audience set based on brand analysis data and the word frequency of the candidate words, and determine the target audience set.

[0181] The targeting categories include at least one of long-tail object categories, attribute object categories, and brand object categories.

[0182] In some embodiments, the target object set determination module may further include:

[0183] The long-tail object class determination module is used to determine that the candidate object belongs to the long-tail object class if the candidate object does not exist in the brand analysis data, and the word frequency of the candidate object is greater than the target number and the word length of the candidate object is greater than or equal to a first preset word length; or,

[0184] The long-tail object category determination module is used to determine that if the candidate object exists in the brand analysis data and the word length of the candidate object is greater than or equal to the first preset word length, then the candidate object belongs to the long-tail object category; wherein, the target quantity is the total number of target products; or,

[0185] The attribute object class determination module is used to determine that if the candidate object exists in the brand analysis data, and the text information of the candidate object contains a preset attribute word and the word length of the candidate object is less than or equal to a second preset word length, then the candidate object belongs to the attribute object class; wherein, the second preset word length is less than the first preset word length; or,

[0186] The brand object class determination module is used to determine that the candidate object belongs to the brand object class if the text information of the candidate object contains the brand information.

[0187] The bidding determination module 250 is used to determine the bidding price for each target target based on the bidding reference range and bidding reference value of each target target in the target target set.

[0188] In some embodiments, the bidding determination module 250 may include:

[0189] The bidding reference range determination module is used to determine the bidding reference range of the target target based on whether there is related historical advertising data for the product in the product line information and the historical advertising data of the online store associated with the product to be advertised.

[0190] In some embodiments, the bidding reference range determination module may include:

[0191] The first bidding reference range determination module is used to determine the bidding reference range corresponding to the target target object if there is associated historical advertising data for the products in the product line information.

[0192] The second bidding reference range determination module is used to determine the bidding reference range corresponding to the target target object if there is no associated historical advertising data for the product in the product line information.

[0193] The third bidding reference range determination module is used to determine the bidding reference range corresponding to the target target object based on the pricing of the product to be advertised, the estimated advertising expenditure ratio, and the estimated conversion rate if the product in the product line information does not have associated historical advertising data and the product in the associated online store of the product to be advertised also does not have historical advertising data.

[0194] The bidding reference value acquisition module is used to acquire the bidding reference value of the target target object;

[0195] The target audience bidding determination module is used to determine the bidding price of the target audience based on the bidding reference range and bidding reference value corresponding to the target audience.

[0196] The creation module 260 is used to create advertisements based on the target audience set, the bid corresponding to the target audience, and the preset budget.

[0197] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described device and module can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0198] In the several embodiments provided in this application, the coupling or direct coupling or communication connection between the modules shown or discussed may be an indirect coupling or communication connection through some interface, device or module, and may be electrical, mechanical or other forms.

[0199] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0200] Please refer to Figure 3. Based on the above-described advertisement creation method, this application embodiment also provides an electronic device 300 that can execute the aforementioned advertisement creation method.

[0201] In embodiments of this application, the electronic device 300 includes one or more processors 310, a memory 320, and one or more application programs. The one or more application programs are stored in the memory 320, which stores programs capable of executing the contents of the foregoing embodiments, and the processor 310 can execute the programs stored in the memory.

[0202] The processor 310 may include one or more cores for data processing and message matrix units. The processor 310 connects to various parts of the electronic device using various interfaces and lines, and performs various functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in memory, and by calling data stored in memory. Optionally, the processor 310 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 310 may integrate one or more of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the displayed content; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 310 and may be implemented separately using a communication chip.

[0203] The memory 320 may include random access memory (RAM) or read-only memory (ROM). The memory 320 can be used to store instructions, programs, code, code sets, or instruction sets. The memory may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function, instructions for implementing the various method embodiments described below, etc. The data storage area may also store data created by the terminal during use.

[0204] Please refer to Figure 4, which shows a schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of this application. The computer-readable medium 400 stores program code, which can be called by a processor to execute the advertising placement adjustment method described in the above method embodiment.

[0205] The computer-readable storage medium 400 may be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. Optionally, the computer-readable storage medium 400 includes a non-transitory computer-readable storage medium. The computer-readable storage medium 400 has storage space for program code 410 that performs any of the method steps described above. This program code can be read from or written to one or more computer program devices. The program code 410 may be compressed, for example, in a suitable form.

[0206] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for creating an advertisement, characterized in that, The method includes: obtaining the target product corresponding to the advertisement to be created; determining the feature information and product line information corresponding to the target product based on a preset database; determining a candidate object set by performing text splitting on the feature information and classifying the search objects corresponding to the product line information; determining a target target object set based on the placement metrics and brand analysis data corresponding to each candidate object in the candidate object set; determining the bid for each target target object based on the bid reference range and bid reference value for each target target object in the target target object set; and creating an advertisement according to the target target object set, the bid corresponding to the target target object, and a preset budget; wherein... The targeting metrics include the word frequency, word length, and brand information of the target product for each candidate object; the word frequency represents the frequency of the candidate object appearing in the candidate object set; the step of determining the target target object set based on the targeting metrics and brand analysis data of each candidate object in the candidate object set includes: classifying the candidate objects based on the brand information of the target product, brand analysis data, word length, word frequency, and text information of the candidate objects to determine the target object set; and sorting the target objects in the target object set based on the brand analysis data and the word frequency of the candidate objects to determine the target target object set.

2. The method according to claim 1, characterized in that, The step of obtaining the target product corresponding to the advertisement to be created includes: obtaining the product to be delivered corresponding to the advertisement to be created; determining the competing products corresponding to the product to be delivered; and determining the target product based on the product to be delivered and the competing products.

3. The method according to claim 2, characterized in that, The step of determining the competing products corresponding to the product to be launched includes: obtaining the category information of the product to be launched; and querying the preset database for products that correspond to the category information and whose sales ranking is within a preset ranking threshold as competing products.

4. The method according to claim 1, characterized in that, The step of determining a candidate object set by performing text segmentation on the feature information and classifying the search objects corresponding to the product line information includes: performing text segmentation on the feature information and assigning the segmented objects to the candidate object set; and classifying the search objects corresponding to the product line information and assigning the classified search objects belonging to the target category to the candidate object set.

5. The method according to claim 4, characterized in that, The feature information includes title information and selling point information; the step of determining the feature information corresponding to the target product based on a preset database includes: determining the title information and selling point information corresponding to the target product based on a preset database.

6. The method according to claim 5, characterized in that, The step of performing text segmentation on the feature information and assigning the segmented objects to a candidate object set includes: splitting the title information and the selling point information into multiple objects with different preset word lengths; and assigning all the segmented objects to a candidate object set.

7. The method according to claim 4, characterized in that, The step of classifying the search objects corresponding to the product line information and assigning the search objects belonging to the target category to the candidate object set includes: classifying the search objects based on the usage benefit index of the search objects corresponding to the product line information; and assigning the search objects belonging to the target category to the candidate object set.

8. The method according to claim 7, characterized in that, The target category includes at least one of high-conversion target category, high-traffic target category, and potential target category; the usage efficiency indicators include: advertising expenditure ratio data, order volume data, and ranking in brand analysis data; the classification of the search object based on the usage efficiency indicators of the search object corresponding to the product line information includes: if the advertising expenditure ratio data corresponding to the search object is less than or equal to a preset advertising expenditure ratio and the corresponding order volume data is greater than or equal to a first preset order volume, then the search object is determined to belong to the high-conversion target category; or, if the advertising expenditure ratio data corresponding to the search object is less than or equal to the preset advertising expenditure ratio, and the corresponding order volume data is greater than a second preset order volume and less than the first preset order volume, then the search object is determined to belong to the potential target category; or, based on brand analysis data in a preset database, the ranking of the search frequency associated with the search object is retrieved; if the ranking of the search object is higher than a preset rank, then the search object is determined to belong to the high-traffic target category.

9. The method according to claim 1, characterized in that, The process of classifying the candidate objects based on the brand information of the target product, brand analysis data, word length, word frequency, and text information of the candidate objects to determine the target object set includes: classifying the candidate objects based on the brand information of the target product, brand analysis data, word length, word frequency, and text information of the candidate objects; and assigning the candidate objects belonging to the targeting category to the target object set.

10. The method according to claim 9, characterized in that, The targeting categories include at least one of long-tail object categories, attribute object categories, and brand object categories. The candidate objects are categorized based on the brand information of the target product, brand analysis data, and the word length, word frequency, and text information of the candidate objects, including: if the candidate object does not exist in the brand analysis data, and the word frequency of the candidate object is greater than the target quantity and the word length of the candidate object is greater than or equal to a first preset word length, then the candidate object is determined to belong to the long-tail object category; or, if the candidate object exists in the brand analysis data, and the word length of the candidate object is greater than or equal to the first preset word length, then the candidate object is determined to belong to the long-tail object category; wherein, the target quantity is the total number of the target products; or, if the candidate object exists in the brand analysis data, and the text information of the candidate object contains preset attribute words and the word length of the candidate object is less than or equal to a second preset word length, then the candidate object is determined to belong to the attribute object category; wherein, the second preset word length is less than the first preset word length; or, if the text information of the candidate object contains the brand information, then the candidate object is determined to belong to the brand object category.

11. The method according to claim 3, characterized in that, The step of determining the bid for each target target based on the bid reference range and bid reference value of each target target in the target target set includes: determining the bid reference range of the target target based on whether there is associated historical advertising data for the product in the product line information and the historical advertising data corresponding to the associated online store of the product to be advertised; obtaining the bid reference value of the target target; and determining the bid for the target target based on the bid reference range and bid reference value corresponding to the target target.

12. The method according to claim 11, characterized in that, The step of determining the bidding reference range for the target audience based on whether there is associated historical advertising data for the products in the product line information and the historical advertising data corresponding to the online stores associated with the products to be advertised includes: if there is associated historical advertising data for the products in the product line information, then the first preset range of average cost-per-click for the corresponding products in the product line information associated with the historical advertising data is used as the bidding reference range for the target audience; if there is no associated historical advertising data for the products in the product line information, then the second preset range of average cost-per-click for the corresponding products in the online stores associated with the products to be advertised that have associated historical advertising data is used as the bidding reference range for the target audience; if there is no associated historical advertising data for the products in the product line information and no historical advertising data for the products in the online stores associated with the products to be advertised, then the bidding reference range for the target audience is determined based on the pricing, estimated advertising expenditure ratio, and estimated conversion rate of the products to be advertised.

13. An advertisement creation device, characterized in that, The device includes: an acquisition module for acquiring the target product corresponding to the advertisement to be created; a first determination module for determining the feature information and product line information corresponding to the target product based on a preset database; a second determination module for determining a candidate object set by performing text splitting on the feature information and classifying the search objects corresponding to the product line information; a third determination module for determining a target advertising object set based on the advertising indicators and brand analysis data corresponding to each candidate object in the candidate object set; a bidding determination module for determining the bidding price of each target advertising object based on the bidding reference range and bidding reference value of each target advertising object in the target advertising object set; and a creation module for creating the advertisement according to the target product information and product line information. The system comprises: a target audience set, bids corresponding to the target audience, and a preset budget for creating advertisements; wherein, the targeting metrics include the keyword frequency, keyword length, and brand information of the target product for each candidate object; the keyword frequency represents the frequency of the candidate object appearing in the candidate object set; the target audience set determination module is used to classify the candidate objects based on the brand information of the target product, brand analysis data, keyword length, keyword frequency, and text information of the candidate objects to determine the target audience set; and the target audience set determination module is used to sort the target objects in the target audience set based on brand analysis data and the keyword frequency of the candidate objects to determine the target audience set.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains program code that can be invoked by a processor to execute the advertisement creation method as described in any one of claims 1-12.

Images