Artificial intelligence-based commodity information intelligent optimization system and method
Patent Information
- Application Number
- CN202611039526.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-14
- Publication Date
- 2026-09-25
AI Technical Summary
一、关键词挖掘不精准、覆盖面窄：商家获取关键词多依赖人工查询平台热搜榜、手动分析竞品Listing标题及描述，不仅耗时耗力，且难以全面捕捉潜在长尾关键词、趋势性关键词；同时，缺乏对关键词热度、竞争度、转化潜力的系统性分析，易导致选择的关键词与商品匹配度低，无法有效提升搜索排名；
1、本发明通过多渠道关键词采集与AI智能匹配技术，整合平台、竞品及类目关键词数据，结合语义分析、属性契合度及用户搜索意图关联等多维度计算，确保关键词与SKU核心属性高度契合，提升Listing搜索相关性与权重，有效提高商品搜索排名与曝光量，从而提升关键词匹配精准度。
Smart Images
Figure CN122817535A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of commodity information processing technology, and in particular to a commodity information intelligent optimization system and method based on artificial intelligence. Background Technology
[0002] In e-commerce platform operations, product listings serve as the core link between merchants and consumers, and their quality directly impacts product search exposure, user clicks, and conversion rates. Currently, platform listing optimization is still primarily manual, with some auxiliary tools only providing basic keyword recommendation functions and lacking end-to-end intelligent support, resulting in numerous prominent technical issues. 1. Inaccurate keyword research and narrow coverage: Merchants often rely on manual searches of trending search lists on platforms and manual analysis of competitor listing titles and descriptions to obtain keywords. This is not only time-consuming and labor-intensive, but also makes it difficult to fully capture potential long-tail keywords and trending keywords. At the same time, the lack of systematic analysis of keyword popularity, competitiveness, and conversion potential can easily lead to low matching between selected keywords and products, making it impossible to effectively improve search rankings. 2. Lack of scientific basis for SKU and keyword matching: When manually matching keywords and SKUs, it relies heavily on the experience and judgment of operations personnel. It is difficult to combine the core attributes of the product, the search habits of the target audience and the platform's algorithm preferences to achieve accurate matching. This often results in keyword stuffing or keywords that are not related to the product's functions and specifications, which affects the listing's search ranking and user experience. Third, the quality of content generation is inconsistent and highly homogenized: When manually writing titles, descriptions and key features, the content is easily limited by professionalism and writing style, resulting in problems such as chaotic content logic, lack of emphasis, and vague expression of selling points; moreover, many listings are imitated by referring to similar competing products, resulting in serious homogenization of listings, making it difficult to stand out among many products and reducing users' willingness to buy; Fourth, low optimization efficiency and poor process integration: Listing optimization requires multiple independent steps such as keyword collection, matching, content writing, manual modification and submission. The process is cumbersome and time-consuming, especially for merchants with multiple SKUs, making it difficult to quickly complete batch optimization. At the same time, when keyword popularity changes or platform rules are adjusted, the listing content needs to be manually adjusted again, which cannot respond to market changes in a timely manner and miss the optimization window. Summary of the Invention
[0003] The purpose of this invention is to solve the problems existing in the prior art by proposing an intelligent optimization system and method for commodity information based on artificial intelligence.
[0004] To achieve the above objectives, the present invention adopts the following technical solution: The AI-based intelligent optimization system for product information includes a keyword collection and analysis module, an AI-SKU intelligent matching module, an AI content generation module, an automated optimization execution module, and an effect monitoring and feedback module. The keyword collection and analysis module serves as a data foundation support unit. It collects keyword data through multiple channels and performs in-depth analysis. The data collection channels of the keyword collection and analysis module include integrated platform interfaces, web crawler technology, and third-party data tools. The collected data includes platform hot search data, competitor keyword data, and category keyword data. The AI-SKU intelligent matching module has a built-in deep learning matching model, and its core function is to achieve accurate matching between keywords and SKUs. The AI content generation module automatically generates high-quality listing content based on platform rules and user needs, and has multiple built-in content generation templates adapted to different categories. The automated optimization execution module, as the core execution unit, interfaces with the platform to enable automatic modification and submission of Listings. The effect monitoring and feedback module is used to monitor the optimization effect in real time and drive secondary optimization.
[0005] As a preferred embodiment, the data processing flow in the keyword collection and analysis module is as follows: Data preprocessing: Clean and deduplicate the collected raw keyword data, and remove invalid and duplicate keywords; Assign weights to metrics: Analyze core metrics such as keyword popularity, competitiveness, conversion potential, and relevance using a preset algorithm; Build a weighted keyword library: Assign corresponding weights to each indicator based on its importance to generate a weighted keyword library.
[0006] As a preferred embodiment, the precise matching process in the AI-SKU intelligent matching module is as follows: Input data: Receive the core attribute information of the SKU to be optimized from the merchant, including key information such as product name, specifications, function, material, and usage scenario; Matching calculation: Call the weighted keyword library, extract the core features of the keywords and the core attributes of the SKU, and perform comprehensive calculations through three dimensions: semantic similarity calculation, attribute fit calculation, and user search intent relevance calculation, and combine the weight of the keywords for weighted processing; Output results: Based on the overall score, the system automatically selects the optimal keyword combination for each SKU, including core keywords that match the product's core attributes, long-tail keywords that cover specific needs, and related keywords that expand search scenarios, ensuring that the keywords are highly relevant to the SKU.
[0007] As a preferred embodiment, the Listing content is generated by the AI content generation module in the following manner: Title generation: An optimized structure combining core keywords, attributes, specifications, and selling points is adopted. Based on the keyword weight, the core keywords are placed at the beginning, the specification and attribute information is inserted in the middle, and long-tail keywords are incorporated at the end to generate a title that meets the platform's length requirements and has high search weight. Description generation: Guided by user needs, it logically and clearly presents the product's core selling points, functional parameters, usage scenarios, and after-sales guarantees, naturally integrating keywords into the content to avoid keyword stuffing and improve content readability and search relevance; Feature highlight generation: Presented in short sentences, it extracts the core competitive advantages of the product, such as cost-effectiveness, unique functions, and quality assurance.
[0008] As a preferred embodiment, the automated optimization execution module implements the automatic modification and submission of Listings as follows: Content compliance verification: The built-in platform listing rule verification engine verifies the title, description, and key features output by the AI content generation module to check for prohibited words, format errors, and excessive length. If the verification fails, an error message is automatically generated and sent to the merchant, or the content is automatically adjusted according to preset rules and re-verified. Differentiation Comparison: Compare the current listing content with the newly generated optimized content, mark the fields that need to be modified, and make it easier for merchants to see and adjust the differences; Automatic modification and submission: Through the platform interface, the title, description, and feature highlight fields in the original listing are automatically replaced, and modification requests are submitted without any manual intervention. Batch operations are also supported, allowing multiple SKUs to be optimized and submitted at once, improving the efficiency of batch optimization.
[0009] As a preferred embodiment, the process by which the effect monitoring and feedback module monitors the optimization effect and drives secondary optimization is as follows: Data monitoring: Real-time tracking of operational data after listing optimization, including search ranking, impressions, click-through rate, conversion rate, and sales metrics; comparison of data changes before and after optimization; and generation of visualized optimization effect reports. Dynamic feedback: Continuously monitor changes in keyword popularity, competitor optimization strategy adjustments, and platform rule updates; when keyword popularity declines, competitor listings are optimized and upgraded, or platform rules change, a secondary optimization reminder is automatically triggered, or a new round of optimization process is initiated based on the conditions preset by the merchant, forming a closed-loop optimization mechanism.
[0010] The AI-based intelligent optimization method for product information includes the following optimization steps: S1: Keyword Collection and Analysis. The keyword collection and analysis module is activated. Through platform interfaces, web crawler technology, and third-party data tools, it synchronously collects trending search data, competitor keyword data, and category keyword data. The collected raw keyword data is cleaned and deduplicated, removing invalid information. Algorithms are used to analyze keyword popularity, competitiveness, conversion potential, and relevance indicators, assigning appropriate weights to each indicator to generate a weighted keyword library, thus completing data preparation. S2: AI-SKU Intelligent Matching. Merchants input the core attribute information of the SKUs to be optimized. The AI-SKU Intelligent Matching module calls a deep learning matching model, reads the weighted keyword library and the core attribute information of the SKU, extracts the core features of the keywords and the core attributes of the SKU, and performs semantic similarity calculation, attribute fit calculation, and user search intent relevance calculation respectively. Combined with the keyword weights, a comprehensive score is ranked, and core keywords, long-tail keywords, and related keywords are selected to generate the optimal keyword combination for each SKU. S3: AI Content Generation. By receiving keyword combinations and SKU core attribute information output by the AI-SKU intelligent matching module, it calls the content generation template adapted to the platform and generates a title according to the structure of core keywords at the beginning, attributes and specifications in the middle, and long-tail keywords at the end, ensuring that the title meets the platform's length requirements. Guided by user needs, it writes descriptions that include core selling points, functional parameters, usage scenarios, and after-sales guarantees, naturally incorporating keywords and extracting the product's core competitive advantages to generate concise and clear feature highlights, thus completing the automatic generation of Listing content. S4: Automated optimization execution. It receives automatically generated Listing content, checks for prohibited words, format errors, and excessive length through a compliance verification engine, and automatically replaces the corresponding fields in the original Listing through the platform interface. It submits a modification request and supports batch operations, allowing multiple SKUs to be modified and submitted at once. S5: Performance Monitoring and Secondary Optimization. Real-time monitoring of operational metrics such as search ranking, exposure, click-through rate, conversion rate, and sales after listing optimization. Comparison of data changes before and after optimization, generation of optimization performance reports. Simultaneously, continuous monitoring of keyword popularity, competitor dynamics, and platform rule changes. When a decrease in keyword popularity, adjustment of competitor optimization strategies, or update of platform rules is detected, a secondary optimization reminder is automatically triggered. If the preset secondary optimization conditions are met, the system automatically starts a new round of optimization process, repeating steps S1-S4 to achieve closed-loop optimization of the listing and ensure continuous and stable optimization results.
[0011] As a preferred option, in step S4, if the verification fails, an error message is generated and fed back to the merchant, or the content is automatically adjusted and re-verified until the verification passes. After the verification passes, the differences between the current Listing content and the newly generated content are compared, and fields that need to be modified are marked.
[0012] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention integrates keyword data from platforms, competitors, and categories through multi-channel keyword collection and AI intelligent matching technology. It combines semantic analysis, attribute fit, and user search intent correlation to ensure that keywords are highly matched with the core attributes of SKUs, improve the relevance and weight of listing search, effectively improve product search ranking and exposure, and thus improve the accuracy of keyword matching.
[0013] 2. The titles, descriptions, and key features generated by AI in this invention strictly adhere to platform rules, align with user reading habits and search algorithm preferences, avoid homogenization issues, enhance product competitiveness, and thus increase user click-through rates and purchase intentions, thereby optimizing the quality of listing content and highlighting differentiation.
[0014] 3. This invention automates the entire process of keyword collection, matching, content generation, modification and submission, eliminating the need for tedious manual operations. It especially supports batch SKU optimization, significantly shortening the optimization cycle. Operations personnel only need to monitor the results and handle abnormal situations, significantly reducing manpower and time costs, thereby improving optimization efficiency and reducing manpower costs.
[0015] 4. This invention tracks operational data and market dynamics in real time through an effect monitoring and feedback module. When keyword popularity changes, platform rules are adjusted, or competitors' strategies change, it triggers secondary optimization in a timely manner to ensure that the listing content is always in the optimal operational state, achieve a steady improvement in conversion rate, thereby dynamically adapting to market changes and continuously improving optimization results. Attached Figure Description
[0016] Figure 1 This is a flowchart illustrating the overall workflow of the AI-based intelligent optimization system for product information proposed in this invention. Figure 2 This is a flowchart of the workflow within the AI-SKU keyword matching module of the intelligent optimization system for product information based on artificial intelligence proposed in this invention. Figure 3 This is a flowchart of the automated listing rule verification process in the intelligent optimization system for product information based on artificial intelligence proposed in this invention. Figure 4 This is a flowchart of the topic generation process within the AI content generation module of the intelligent optimization system for commodity information based on artificial intelligence proposed in this invention.
[0017] Figure 5 This is a flowchart of the intelligent optimization method for commodity information based on artificial intelligence proposed in this invention. Detailed Implementation
[0018] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0019] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.
[0020] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.
[0021] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.
[0022] Example, refer to Figures 1 to 4 The AI-based intelligent optimization system for product information includes a keyword collection and analysis module, an AI-SKU intelligent matching module, an AI content generation module, an automated optimization execution module, and an effect monitoring and feedback module. The keyword collection and analysis module serves as a data foundation support unit. It collects keyword data from multiple channels and performs in-depth analysis. The data collection channels for the keyword collection and analysis module are integrated platform interfaces, web crawler technology, and third-party data tools. The collected data includes platform hot search data, competitor keyword data, and category keyword data. Furthermore, the data processing flow in the keyword collection and analysis module is as follows: Data preprocessing: Clean and deduplicate the collected raw keyword data, and remove invalid and duplicate keywords; Assign weights to metrics: Analyze core metrics such as keyword popularity, competitiveness, conversion potential, and relevance using a preset algorithm; Build a weighted keyword library: Assign corresponding weights to each indicator based on its importance to generate a weighted keyword library.
[0023] When conducting multi-dimensional keyword indicator analysis, the following comprehensive keyword weight formula is often used:
[0024] in, This is the overall keyword weight (value range [0,1]), used to filter high-quality keywords; The keyword popularity value (range [0,1]) represents the popularity of keyword searches; The keyword conversion potential (value range [0,1]) represents the probability that a keyword will lead to a purchase; Keyword relevance (value range [0,1]) represents the degree of fit between keywords and category / product attributes; Keyword competitiveness (value range [0,1]), representing the density of keyword usage (negative indicator); and The index weight coefficient, and ,and and The default initial values are obtained through training using historical data from the platform. , ,and and It can be dynamically adjusted according to category characteristics; Based on the above, The data is collected by the platform's search ranking, the frequency of related recommended keywords, and the search volume of long-tail keywords; or by the platform interface or web crawler to count the keyword search volume in the past 7-30 days. After normalization, the value range is [0,1]. By counting the number of competing listings using the keyword in the target category and analyzing the search ranking distribution of listings using the keyword, the competition intensity coefficient is calculated and the value range is [0,1] after normalization. By analyzing the matching degree between keywords and the core attributes of the category (such as the correlation between the category "sports shoes" and the keyword "waterproof"), and combining the preliminary screening results of the core attributes of the SKU, the fit degree is calculated by semantic matching algorithm, with a value range of [0,1].
[0025] The AI-SKU intelligent matching module has a built-in deep learning matching model, and its core function is to achieve accurate matching between keywords and SKUs. Furthermore, the precise matching process in the AI-SKU intelligent matching module is as follows: Input data: Receive the core attribute information of the SKU to be optimized from the merchant, including key information such as product name, specifications, function, material, and usage scenario; Matching calculation: Call the weighted keyword library, extract the core features of the keywords and the core attributes of the SKU, and perform comprehensive calculations through three dimensions: semantic similarity calculation, attribute fit calculation, and user search intent relevance calculation, and combine the weight of the keywords for weighted processing; Output results: Based on the overall score, the best keyword combination is automatically selected for each SKU, including core keywords that match the core attributes of the product, long-tail keywords that cover segmented needs, and related keywords that expand the search scenarios, ensuring that the keywords are highly consistent with the SKU; The formula for calculating semantic similarity score is as follows:
[0026] in, The semantic similarity score (value range [0,1]) is given, with higher values indicating stronger semantic fit. The core feature vector of the keyword (an n-dimensional vector, where n is a fixed dimension of the NLP model, such as 768 dimensions in the BERT model). The feature vector of the core attribute information of the SKU (and) (n-dimensional vectors of the same dimension) and for and The vector magnitude; Based on the above, Using NLP pre-trained models such as BERT, keywords are segmented and semantically encoded to generate fixed-dimensional vectors; Adopted and The same NLP pre-trained model is used to segment and semantically encode the SKU attribute text entered by the merchant.
[0027] The formula for calculating the attribute fit score is as follows:
[0028] in, The attribute fit score is calculated (value range [0,1]), with higher values indicating more accurate attribute matching. This refers to the total number of core attribute tags for the SKU, such as the number of attribute categories like material, function, and specifications. This is the attribute matching identifier (1 is set when the i-th SKU attribute tag matches the keyword attribute tag, and 0 is set when they do not match). The weight of the i-th attribute tag is assigned based on the importance of the attribute to the product, such as a weight of 0.3 for core functions and 0.2 for material, etc., satisfying the following conditions: ; The formula for calculating the search intent relevance score is as follows:
[0029] in, The search intent relevance score (value range [0,1]) is given; the higher the value, the better it matches the user's purchase needs. For direct matching, the value is 1 if the keyword search intent is exactly the same as the SKU requirement tag, and 0 otherwise. As an indirect association identifier, it is set to 1 if the keyword search intent is related to but not completely consistent with the SKU demand tag, and 0 otherwise; and For the correlation weight coefficient, satisfying Default initial value: ; Based on the above, Based on platform search log data, purchase intent tags corresponding to keywords are classified and identified through a neural network model. By parsing the usage scenario descriptions in the core attributes of the SKU, they are mapped to the system's preset requirement tag library.
[0030] In the process of calculating the overall score, the formula for calculating the overall score is as follows:
[0031] in, The overall score (range [0,1]) is the core criterion for keyword selection. Sort in descending order to filter the top 1-3 core keywords, top 4-13 long-tail keywords, and top 14-17 related keywords; and These are scores for semantic similarity, attribute fit, and search intent relevance, respectively. The overall weight of keywords; and For the dimension weight coefficients, satisfying Calculated by training on historical data matching platform SKUs, with default initial values: ; Based on the above, To measure the contribution of keywords to whether their semantic connotations are consistent with the semantics of the core attributes of the SKU; To measure the contribution of keywords in accurately matching the specific attributes of the SKU; To measure the contribution of keywords to whether the user search intent behind them matches the user needs that the SKU can meet; This measures the contribution of the quality of the keywords themselves (popularity, conversion potential, etc.) to the matching effect.
[0032] The AI content generation module automatically generates high-quality listing content based on platform rules and user needs, and has multiple built-in content generation templates adapted to different categories. Furthermore, the Listing content generation method of the AI content generation module is as follows: Title generation: An optimized structure combining core keywords, attributes, specifications, and selling points is adopted. Based on the keyword weight, the core keywords are placed at the beginning, the specification and attribute information is inserted in the middle, and long-tail keywords are incorporated at the end to generate a title that meets the platform's length requirements and has high search weight. Description generation: Guided by user needs, it logically and clearly presents the product's core selling points, functional parameters, usage scenarios, and after-sales guarantees, naturally integrating keywords into the content to avoid keyword stuffing and improve content readability and search relevance; Feature highlight generation: Presented in short sentences, it extracts the core competitive advantages of the product, such as cost-effectiveness, unique functions, and quality assurance.
[0033] The formula for calculating the keyword position ranking score is as follows:
[0034] in, The keyword position is ranked and scored, with higher values indicating higher priority. For keyword comprehensive weight, This is a positional weighting factor for the first 30% of the title text. 40% of the character area in the middle 30% of the character area is back ; The verification formula for title length compliance checking and simplification calculation is as follows:
[0035] in, To optimize the number of characters in the compliant title; This represents the initial number of characters in the title before simplification. This is the platform title length threshold extracted from platform rules.
[0036] The automated optimization and execution module serves as the core execution unit, connecting to the platform interface to enable automatic modification and submission of listings. Based on the above, This is obtained by counting the number of characters (including spaces and punctuation) in the initial title. Extracted from the platform's rules document, updated in real-time via the platform's API, and then integrated into the system. like Keep it directly; if Simplify according to the priority of "core keywords > attributes > specifications > selling points" to ensure... .
[0037] Furthermore, the automated optimization execution module implements the automatic modification and submission of Listings as follows: Content compliance verification: The built-in platform listing rule verification engine verifies the title, description, and key features output by the AI content generation module to check for prohibited words, format errors, and excessive length. If the verification fails, an error message is automatically generated and sent to the merchant, or the content is automatically adjusted according to preset rules and re-verified. Differentiation Comparison: Compare the current listing content with the newly generated optimized content, mark the fields that need to be modified, and make it easier for merchants to see and adjust the differences; Automatic modification and submission: Through the platform interface, the title, description, and feature highlight fields in the original listing are automatically replaced, and modification requests are submitted without any manual intervention. Batch operations are also supported, allowing multiple SKUs to be optimized and submitted at once, improving the efficiency of batch optimization.
[0038] The set formula for forbidden word detection and replacement calculation is as follows:
[0039] in, To optimize the compliant vocabulary set; This is the initial vocabulary set for titles after word segmentation. This is a set of prohibited words (the intersection of the initial vocabulary and the prohibited word list). Replace the vocabulary set for compliance (and) (compliant words with semantic similarity ≥ 0.8). The platform's banned word library (built-in static library + real-time updated library). Words prohibited by the platform (including false advertising terms, illegal expressions, and infringing terms); Based on the above, We collect data from platform interfaces and official rule documents, and regularly update and integrate it into the system. By performing Chinese / English word segmentation on the initial title, a vocabulary set is obtained; Identification is achieved through set intersection operations; Match words with a semantic similarity of ≥0.8 from the compliant keyword database; if no match is found, delete the non-compliant words and add synonymous compliant expressions.
[0040] The performance monitoring and feedback module is used to monitor the optimization effect in real time and drive secondary optimization; Furthermore, the process by which the effect monitoring and feedback module monitors the optimization effect and drives secondary optimization is as follows: Data monitoring: Real-time tracking of operational data after listing optimization, including search ranking, impressions, click-through rate, conversion rate, and sales metrics; comparison of data changes before and after optimization; and generation of visualized optimization effect reports. Dynamic feedback: Continuously monitor changes in keyword popularity, competitor optimization strategy adjustments, and platform rule updates; when keyword popularity declines, competitor listings are optimized and upgraded, or platform rules change, a secondary optimization reminder is automatically triggered, or a new round of optimization process is initiated based on the conditions preset by the merchant, forming a closed-loop optimization mechanism.
[0041] The weighting coefficients of all formulas are iteratively optimized using operational data from the performance monitoring feedback module, such as search ranking and conversion rate. The iterative optimization formulas are as follows:
[0042] in, These are the coefficients after iteration; For the current coefficient, For learning rate, To optimize the objective function ( This directly links to the three core objectives that merchants care about most: search ranking, user purchase conversion, and product exposure, thus avoiding weight adjustments that deviate from actual business needs. Based on the above, Sensitivity to the impact of the current weights on the optimization effect is a quantification of the impact of a one-unit change in the current weights on the optimization effect of the objective function. The magnitude and direction of the change, when This indicates that increasing the current weight can improve the optimization effect. ,when This indicates that increasing the current weight will lead to an improvement in the optimization effect. Decline, and The larger the absolute value, the greater the current weight. The more significant the impact.
[0043] Reference Figure 5 The AI-based intelligent optimization method for product information includes the following optimization steps: S1: Keyword Collection and Analysis. The keyword collection and analysis module is activated. Through platform interfaces, web crawler technology, and third-party data tools, it synchronously collects trending search data, competitor keyword data, and category keyword data. The collected raw keyword data is cleaned and deduplicated, removing invalid information. Algorithms are used to analyze keyword popularity, competitiveness, conversion potential, and relevance indicators, assigning appropriate weights to each indicator to generate a weighted keyword library, thus completing data preparation. S2: AI-SKU Intelligent Matching. Merchants input the core attribute information of the SKUs to be optimized. The AI-SKU Intelligent Matching module calls a deep learning matching model, reads the weighted keyword library and the core attribute information of the SKU, extracts the core features of the keywords and the core attributes of the SKU, and performs semantic similarity calculation, attribute fit calculation, and user search intent relevance calculation respectively. Combined with the keyword weights, a comprehensive score is ranked, and core keywords, long-tail keywords, and related keywords are selected to generate the optimal keyword combination for each SKU. S3: AI Content Generation. By receiving keyword combinations and SKU core attribute information output by the AI-SKU intelligent matching module, it calls the content generation template adapted to the platform and generates a title according to the structure of core keywords at the beginning, attributes and specifications in the middle, and long-tail keywords at the end, ensuring that the title meets the platform's length requirements. Guided by user needs, it writes descriptions that include core selling points, functional parameters, usage scenarios, and after-sales guarantees, naturally incorporating keywords and extracting the product's core competitive advantages to generate concise and clear feature highlights, thus completing the automatic generation of Listing content. S4: Automated optimization execution. It receives automatically generated Listing content, checks for prohibited words, format errors, and excessive length through a compliance verification engine, and automatically replaces the corresponding fields in the original Listing through the platform interface. It submits a modification request and supports batch operations, allowing multiple SKUs to be modified and submitted at once. Furthermore, in step S4, if the verification fails, an error message is generated and fed back to the merchant, or the content is automatically adjusted and re-verified until the verification passes. After the verification passes, the differences between the current Listing content and the newly generated content are compared, and fields that need to be modified are marked.
[0044] S5: Performance Monitoring and Secondary Optimization. Real-time monitoring of operational metrics such as search ranking, exposure, click-through rate, conversion rate, and sales after listing optimization. Comparison of data changes before and after optimization, generation of optimization performance reports. Simultaneously, continuous monitoring of keyword popularity, competitor dynamics, and platform rule changes. When a decrease in keyword popularity, adjustment of competitor optimization strategies, or update of platform rules is detected, a secondary optimization reminder is automatically triggered. If the preset secondary optimization conditions are met, the system automatically starts a new round of optimization process, repeating steps S1-S4 to achieve closed-loop optimization of the listing and ensure continuous and stable optimization results.
[0045] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. An intelligent optimization system for commodity information based on artificial intelligence, characterized in that: It includes a keyword collection and analysis module, an AI-SKU intelligent matching module, an AI content generation module, an automated optimization execution module, and an effect monitoring and feedback module; The keyword collection and analysis module serves as a data foundation support unit. It collects keyword data through multiple channels and performs in-depth analysis. The data collection channels of the keyword collection and analysis module include integrated platform interfaces, web crawler technology, and third-party data tools. The collected data includes platform hot search data, competitor keyword data, and category keyword data. The AI-SKU intelligent matching module has a built-in deep learning matching model, and its core function is to achieve accurate matching between keywords and SKUs. The AI content generation module automatically generates high-quality listing content based on platform rules and user needs, and has multiple built-in content generation templates adapted to different categories. The automated optimization execution module, as the core execution unit, interfaces with the platform to enable automatic modification and submission of Listings. The effect monitoring and feedback module is used to monitor the optimization effect in real time and drive secondary optimization.
2. The intelligent optimization system for commodity information based on artificial intelligence according to claim 1, characterized in that, The data processing flow in the keyword collection and analysis module is as follows: Data preprocessing: Clean and deduplicate the collected raw keyword data, and remove invalid and duplicate keywords; Assign weights to metrics: Analyze core metrics such as keyword popularity, competitiveness, conversion potential, and relevance using a preset algorithm; Build a weighted keyword library: Assign corresponding weights to each indicator based on its importance to generate a weighted keyword library.
3. The intelligent optimization system for commodity information based on artificial intelligence according to claim 1, characterized in that, The precise matching process in the AI-SKU intelligent matching module is as follows: Input data: Receive the core attribute information of the SKU to be optimized from the merchant, including key information such as product name, specifications, function, material, and usage scenario; Matching calculation: Call the weighted keyword library, extract the core features of the keywords and the core attributes of the SKU, and perform comprehensive calculations through three dimensions: semantic similarity calculation, attribute fit calculation, and user search intent relevance calculation, and combine the weight of the keywords for weighted processing; Output results: Based on the overall score, the system automatically selects the optimal keyword combination for each SKU, including core keywords that match the product's core attributes, long-tail keywords that cover specific needs, and related keywords that expand search scenarios, ensuring that the keywords are highly relevant to the SKU.
4. The intelligent optimization system for commodity information based on artificial intelligence according to claim 1, characterized in that, The Listing content is generated in the manner described in the AI content generation module as follows: Title generation: An optimized structure combining core keywords, attributes, specifications, and selling points is adopted. Based on the keyword weight, the core keywords are placed at the beginning, the specification and attribute information is inserted in the middle, and long-tail keywords are incorporated at the end to generate a title that meets the platform's length requirements and has high search weight. Description generation: Guided by user needs, it logically and clearly presents the product's core selling points, functional parameters, usage scenarios, and after-sales guarantees, naturally integrating keywords into the content to avoid keyword stuffing and improve content readability and search relevance; Feature highlight generation: Presented in short sentences, it extracts the core competitive advantages of the product, such as cost-effectiveness, unique functions, and quality assurance.
5. The intelligent optimization system for commodity information based on artificial intelligence according to claim 1, characterized in that, The automated optimization execution module implements the automatic modification and submission of Listings as follows: Content compliance verification: The built-in platform listing rule verification engine verifies the title, description, and feature highlights output by the AI content generation module to check for violations such as prohibited words, format errors, and excessive length. If the verification fails, an error message will be automatically generated and sent to the merchant, or the content will be automatically adjusted according to preset rules and the verification will be repeated. Differentiation Comparison: Compare the current listing content with the newly generated optimized content, mark the fields that need to be modified, and make it easier for merchants to see and adjust the differences; Automatic modification and submission: Through the platform interface, the title, description, and feature highlight fields in the original listing are automatically replaced, and modification requests are submitted without any manual intervention. Batch operations are also supported, allowing multiple SKUs to be optimized and submitted at once, improving the efficiency of batch optimization.
6. The intelligent optimization system for commodity information based on artificial intelligence according to claim 1, characterized in that, The process by which the effect monitoring and feedback module monitors the optimization effect and drives secondary optimization is as follows: Data monitoring: Real-time tracking of operational data after listing optimization, including search ranking, impressions, click-through rate, conversion rate, and sales metrics; comparison of data changes before and after optimization; and generation of visualized optimization effect reports. Dynamic feedback: Continuously monitor changes in keyword popularity, competitor optimization strategy adjustments, and platform rule updates; when keyword popularity declines, competitor listings are optimized and upgraded, or platform rules change, a secondary optimization reminder is automatically triggered, or a new round of optimization process is initiated based on the conditions preset by the merchant, forming a closed-loop optimization mechanism.
7. An AI-based intelligent optimization method for commodity information, applied to the AI-based intelligent optimization system for commodity information according to claims 1-6, characterized in that, Includes the following steps: S1: Keyword Collection and Analysis. The keyword collection and analysis module is activated. Through platform interfaces, web crawler technology, and third-party data tools, it synchronously collects platform hot search data, competitor keyword data, and category keyword data. The collected raw keyword data is cleaned, deduplicated, and invalid information is removed. By analyzing the popularity, competitiveness, conversion potential, and relevance of keywords using algorithms, and assigning corresponding weights to each indicator, a weighted keyword library is generated, thus completing the data preparation. S2: AI-SKU Intelligent Matching. Merchants input the core attribute information of the SKUs to be optimized. The AI-SKU Intelligent Matching module calls a deep learning matching model, reads the weighted keyword library and the core attribute information of the SKUs, extracts the core features of the keywords and the core attributes of the SKUs, and performs semantic similarity calculation, attribute fit calculation, and user search intent relevance calculation respectively. Combined with the keyword weights, a comprehensive score is ranked, and core keywords, long-tail keywords, and related keywords are selected to generate the optimal keyword combination for each SKU. S3: AI Content Generation. By receiving keyword combinations and SKU core attribute information output by the AI-SKU intelligent matching module, it calls the content generation template adapted to the platform and generates a title according to the structure of core keywords at the beginning, attributes and specifications in the middle, and long-tail keywords at the end, ensuring that the title meets the platform's length requirements. Guided by user needs, it writes descriptions that include core selling points, functional parameters, usage scenarios, and after-sales guarantees, naturally incorporating keywords and extracting the product's core competitive advantages to generate concise and clear feature highlights, thus completing the automatic generation of Listing content. S4: Automated optimization execution. It receives automatically generated Listing content, checks for prohibited words, format errors, and excessive length through a compliance verification engine, and automatically replaces the corresponding fields in the original Listing through the platform interface. It submits a modification request and supports batch operations, allowing multiple SKUs to be modified and submitted at once. S5: Performance Monitoring and Secondary Optimization. Real-time monitoring of operational metrics such as search ranking, exposure, click-through rate, conversion rate, and sales after listing optimization. Comparison of data changes before and after optimization, generation of optimization performance reports. Simultaneously, continuous monitoring of keyword popularity, competitor dynamics, and platform rule changes. When a decrease in keyword popularity, adjustment of competitor optimization strategies, or update of platform rules is detected, a secondary optimization reminder is automatically triggered. If the preset secondary optimization conditions are met, the system automatically starts a new round of optimization process, repeating steps S1-S4 to achieve closed-loop optimization of the listing and ensure continuous and stable optimization results.
8. The intelligent optimization method for commodity information based on artificial intelligence according to claim 7, characterized in that, In step S4, if the verification fails, an error message is generated and fed back to the merchant, or the content is automatically adjusted and re-verified until the verification passes. After the verification passes, the differences between the current Listing content and the newly generated content are compared, and fields that need to be modified are marked.