A method and system for internet content management in the footwear industry

Abstract

This invention provides a method and system for internet content management in the footwear industry. The method includes acquiring structured product data and digital asset data of footwear from e-commerce platforms, brand content libraries, and social media; parsing, standardizing, and deduplicating content entries; calculating multiple sub-scores for deduplicated content entries based on footwear field completeness, size system consistency, attribute constraint consistency, digital asset-master data matching degree, and channel adaptability; generating quality scores through weight fusion and constraint conflict correction; and triggering a workflow status for publishable, incomplete, or reviewed entries based on the scores; performing compliance review using a rule engine on publishable or reviewed entries, and generating and storing a publication snapshot. This solution achieves automated quality control and intelligent publishing of content entries through multi-dimensional algorithms and dynamic weight adjustments, improving content accuracy, reducing manual review, optimizing multi-channel publishing efficiency, and enhancing data traceability.

Description

Technical Field

[0001] This invention belongs to the field of internet content technology for the footwear industry, and specifically relates to a method and system for internet content management in the footwear industry. Background Technology

[0002] With the rapid development of e-commerce platforms and social media, footwear companies face a large amount of product information and digital assets from different channels. Existing content management methods usually rely on manual input or simple rule matching, making it difficult to achieve automatic alignment and efficient management of cross-platform data. At the same time, traditional methods lack unified quantitative evaluation means for product attribute completeness, size system consistency, image and text matching degree, and channel adaptability, leading to problems such as duplication, incomplete information, or image and text discrepancies within products, which in turn increases the burden of manual review and the risk of publishing errors. Therefore, a technical solution is needed that can automatically deduplicate, evaluate the quality, and intelligently publish content items from multiple sources to improve the accuracy and publishing efficiency of footwear content items.

[0003] Existing internet content management solutions for the footwear industry, such as the patent "An Internet Website Content Management System" (publication number: CN102609526A), can achieve unified management, review, and publication of content from multiple websites. However, its technical focus is mainly on website content updates, workflow control, and webpage publishing. It lacks sufficient processing for footwear industry content items, such as footwear master data and digital asset association modeling, multi-source content similarity deduplication, quality scoring based on size system and attribute constraints, and publication snapshot retention. Therefore, its adaptability to the refined governance of product content and multi-channel publication quality control in the footwear industry internet scenario is still limited. Summary of the Invention

[0004] To address the aforementioned issues, this application provides a method for internet content management in the footwear industry.

[0005] Includes the following steps:

[0006] S1. Obtain shoe industry-related content data from two heterogeneous data sources. The content data includes structured product data and unstructured digital asset data, and add a source identifier and timestamp to each piece of content data.

[0007] S2. Based on a preset footwear content data model, the content data is parsed and standardized to generate content entries; wherein, the content entries include: a footwear master data domain and a digital asset domain.

[0008] The footwear master data field contains footwear identifiers and a set of footwear attributes, and the digital asset field contains digital asset references associated with the footwear identifiers;

[0009] S3. Based on the conflict resolution rules calculated by similarity, perform deduplication on the content entries;

[0010] S4. Perform a quality assessment on the deduplicated content items to obtain a quality score, and trigger a workflow status change for the content items based on the quality score. Step S4 includes:

[0011] S41. Extract the footwear master data features, digital asset features, and channel release parameters of the content entries;

[0012] S42. Based on the footwear master data characteristics, digital asset characteristics, and channel release parameters, calculate multiple sub-scores respectively. The sub-scores include at least the footwear field completeness score, size system consistency score, footwear attribute constraint consistency score, digital asset and footwear master data matching score, and channel adaptation score.

[0013] S43. Merge the sub-scores according to the preset weighting rules to obtain the basic quality score;

[0014] S44. Based on the constraint conflicts caused by missing key fields, size system conflicts, mismatch between digital assets and footwear master data, and hard non-compliance with channel rules, the basic quality score is corrected to obtain the quality score.

[0015] S45. The quality score is compared with the preset publishing threshold and the completion threshold respectively to trigger a change in the workflow status of the content item. The workflow status of the content item includes a publishable status and a completion-required status.

[0016] S5. Perform compliance review on content items that are in the publishable or review-required state. The compliance review includes rule engine review. Publish the content item after the review is passed, and record the review evidence and handling results when the review fails.

[0017] S6. Generate and store a release snapshot each time a release is made. The release snapshot includes the channel release package, content entry version number, operator identifier, sending time and channel receipt, and stores the release snapshot to a storage medium that meets the preset retention period.

[0018] Preferably, the heterogeneous data sources include content from e-commerce platforms, brand-owned content sources, and social media content.

[0019] Preferably, the footwear attribute set includes category, gender, size system, color, material, and applicable scenarios.

[0020] Preferably, the conflict resolution rules include: retaining data according to source priority, retaining data according to the latest timestamp, retaining data according to manual confirmation marks, and merging and retaining content items from multiple sources based on shoe identification.

[0021] Preferably, the completeness score of the footwear field is obtained in the following manner:

[0022] The importance coefficient of each field is determined according to the preset set of required fields, and the completeness score of the footwear field is calculated by the ratio of the sum of the importance coefficients of the filled and format-validated fields to the sum of the total importance coefficients of the preset set of required fields.

[0023] Preferably, the size system consistency score is obtained by matching the size data in the content entries with a preset size mapping library; the footwear attribute constraint consistency score is obtained by verifying the preset constraint relationship between category, gender, size system and applicable scenario.

[0024] Preferably, the rule engine review process includes: the matching score between digital assets and footwear master data is obtained by extracting the visual or textual features of the digital assets and performing similarity matching with the title, description, or attribute tags in the footwear master data domain; the channel adaptation score is obtained by verifying the title length, image specifications, category mapping, and compliance of prohibited word rules of the target channel.

[0025] Preferably, the preset weight rules and the release threshold and completion threshold are dynamically adjusted based on historical manual review results, post-release feedback results or error correction records. The rule engine review process includes prohibited word matching, channel source blacklist matching, grammatical structure legality determination and trademark or brand field compliance verification.

[0026] Preferably, the rule engine review process includes: matching prohibited words, matching the blacklist of channel sources, and determining the legality of the grammatical structure.

[0027] This invention also provides an internet content management system for the footwear industry, applicable to the above-mentioned method, comprising:

[0028] The data access module is used to obtain footwear industry-related content data from heterogeneous data sources and attach source identifiers, timestamps, and tenant identifiers.

[0029] The content modeling module is used to generate content entries based on the footwear content data model.

[0030] The alignment and deduplication module is used to deduplicate content items;

[0031] The Quality Assessment and Workflow module is used to perform quality assessments on deduplicated content items, generate quality scores, and automatically set the workflow status of content items.

[0032] The compliance review module is used to perform rule engine reviews on content items and record review evidence and processing results;

[0033] The publishing module is used to publish content items, and at the same time generate a publishing snapshot and store it according to a preset retention period;

[0034] The quality assessment and workflow module includes:

[0035] The feature extraction unit is used to extract footwear master data features, digital asset features, and channel release parameters;

[0036] Sub-scoring calculation units are used to calculate the completeness score of footwear fields, the consistency score of the size system, the consistency score of footwear attribute constraints, the matching score between digital assets and footwear master data, and the channel adaptation score, respectively.

[0037] The fusion scoring unit is used to fuse the sub-scorings according to preset weighting rules to obtain a basic quality score;

[0038] The scoring correction unit is used to correct the basic quality score based on the constraint conflict terms to obtain a quality score;

[0039] The status switching unit is used to compare the quality score with preset publishing thresholds and completion thresholds, and trigger a workflow status change for the content item.

[0040] The present invention has the following beneficial effects:

[0041] This invention provides a method and system for internet content management in the footwear industry. The method includes acquiring structured product data and digital asset data of footwear from e-commerce platforms, brand content libraries, and social media; parsing, standardizing, and deduplicating content entries; calculating multiple sub-scores for deduplicated content entries based on footwear field completeness, size system consistency, attribute constraint consistency, digital asset-master data matching degree, and channel adaptability; generating quality scores through weight fusion and constraint conflict correction; and triggering a workflow status for publishable, incomplete, or reviewed entries based on the scores; performing compliance review using a rule engine on publishable or reviewed entries, and generating and storing a publication snapshot. This solution achieves automated quality control and intelligent publishing of content entries through multi-dimensional algorithms and dynamic weight adjustments, improving content accuracy, reducing manual review, optimizing multi-channel publishing efficiency, and enhancing data traceability. Attached Figure Description

[0042] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0043] Figure 1 This is a system workflow diagram of the present invention;

[0044] Figure 2This is a system module block diagram of the present invention;

[0045] Figure 3 This is a block diagram of the quality assessment and workflow module of the present invention. Detailed Implementation

[0046] Example 1

[0047] See Figure 1 This application provides a method for managing internet content in the footwear industry, which includes the following steps:

[0048] Footwear industry-related content data is acquired from two heterogeneous data sources. This content data includes structured product data and unstructured digital asset data. Each piece of content data is appended with a source identifier, timestamp, and tenant identifier. Heterogeneous data sources refer to data providers whose data formats, protocols, levels of structuring, or update mechanisms differ. Structured product data refers to footwear product metadata with clearly defined fields and constraints, including shoe style identifiers, categories, gender, size systems, colors, materials, prices, inventory, brands, and applicable scenarios. Unstructured digital asset data refers to media resources carrying visual, auditory, or interactive semantics, including shoe images, short videos, or 3D models. The source identifier uniquely identifies the data provider to distinguish the authority and credibility levels of different sources and provides a basis for conflict resolution and traceability audits. The timestamp is accurate to milliseconds and is used for sorting, version management, and deduplication. The tenant identifier distinguishes content data from different brands or business units.

[0049] For example, this application could involve a domestic sports shoe brand accessing the Taobao / Tmall merchant backend, Xiaohongshu brand account API, and an Excel product list exported from its internal PLM system. Within a single synchronization cycle, it could acquire structured product data and unstructured digital assets. All data would be automatically appended with corresponding source identifiers, timestamps, and tenant identifiers before being stored in the database. This would ensure that subsequent content item deduplication, quality assessment, and workflow status switching could be executed accurately, and that the content data source would be traceable, the version manageable, and the heterogeneous data compatible.

[0050] Step Two: Based on a pre-defined footwear content data model, the acquired content data is parsed and standardized to generate content entries. Each content entry includes at least a footwear master data domain and a digital asset domain. The footwear master data domain contains footwear identifiers and a set of footwear attributes, while the digital asset domain contains one or more digital asset references associated with the footwear identifiers. The footwear content data model is a domain ontology model oriented towards semantic abstraction of footwear industry business. It defines a standardized description framework with footwear as the core entity, used to uniformly describe footwear attributes and their business relationships. The footwear identifier is a globally unique and business-readable string, serving as a reference between the master data domain and the digital asset domain. Anchor points between digital asset domains ensure that footwear attributes and digital assets logically belong to the same footwear entity. The footwear attribute set is a structured group of fields organized according to business dimensions. Its function is to carry computable semantics such as the footwear's function, appearance, size system, color, material, and applicable scenarios, and to support the calculation of various sub-scores in subsequent quality assessments. The digital asset reference is a lightweight handle that points to the storage location instead of directly storing the original file. Its function is to achieve decoupling management and on-demand loading of master data and massive unstructured resources, thereby ensuring fast access and consistent maintenance of content items in a multi-channel, high-concurrency environment.

[0051] For example, this application can map SKU data from Taobao to shoe style identifiers, parse their fields and fill them into the corresponding fields in the shoe category attribute set, and bind the short video URL returned by the Xiaohongshu API and the 3D model OSS Key exported by the PLM system as digital asset references to the same shoe style identifier, forming a complete and standardized content entry, thereby providing a unified and computable data foundation for subsequent deduplication, quality assessment and workflow status triggering.

[0052] Step 3: Deduplicate content entries based on conflict resolution rules. The conflict resolution rules are a set of decision logic used to determine whether multiple content entries point to the same shoe entity and to determine the basis for retention. Their function is to avoid the expansion of shoe master data and the confusion of digital asset references caused by multiple source duplicate submissions, version mismatches or parsing ambiguities, thereby ensuring the uniqueness of content entries and data consistency.

[0053] For example, the system first performs precise matching based on shoe brand identifiers, and entries with identical identifiers enter the conflict detection process. Subsequently, the system calculates the semantic similarity of content entries on the dimension of shoe attribute set, including the similarity of fields such as category, gender, size system, color, material, and applicable scenarios, and performs a comprehensive similarity assessment in combination with the visual or textual features of digital assets. When the similarity is higher than a preset threshold, the entry is identified as a potential conflict. Finally, the system selects the best entry to retain according to a preset priority order (such as source credibility, update time, and manual confirmation mark), and marks the remaining entries as deduplicated and archives them for future reference, so as to be used for subsequent traceability, quality assessment, and workflow processing, while avoiding duplicate data from interfering with scoring and publication.

[0054] Step Four: Perform a quality assessment on the deduplicated content items to obtain a quality score, and trigger a workflow status change for the content items based on the quality score. This step aims to comprehensively determine the usability of content items in terms of data completeness, business consistency, resource matching, and channel publishability, thus forming a unified quantitative result for the usability of content items before entering the subsequent compliance review and release process. To this end, the system first extracts footwear master data features, digital asset features, and channel release parameters from the content items. Footwear master data features include at least category, gender, size system, color, material, applicable scenarios, title, and description; digital asset features include at least image subject features, video cover features, resource size information, or text information associated with the resource; and channel release parameters include at least the target channel's title length limit, image specification requirements, category mapping rules, and prohibited word rules.

[0055] The quality score is a quantitative indicator that comprehensively assesses the performance of content items across multiple evaluation dimensions. These dimensions include at least the completeness of footwear fields, consistency of the size system, consistency of footwear attribute constraints, matching degree between digital assets and footwear master data, and channel adaptability. Footwear field completeness reflects the effective filling degree of required fields in the content item. The system determines the importance coefficient of each field based on a preset set of required fields and calculates the footwear field completeness score by the ratio of the sum of the importance coefficients of filled and format-validated fields to the sum of the total importance coefficients of the preset set of required fields. Size system consistency reflects the overall performance of the content item across multiple evaluation dimensions. The matching degree between the size data in the entry and the preset size mapping library is used to avoid situations such as size range mismatch, mixing of men's and women's shoe sizes, or incorrect regional size mapping; the consistency of footwear attribute constraints is used to reflect whether the preset business constraint relationship is met between the category, gender, size system, and applicable scenario to avoid attribute logic conflicts; the matching degree between digital assets and footwear master data is used to reflect the degree of correspondence between images, videos, or other digital resources and the titles, descriptions, or attribute tags in the footwear master data field; the channel adaptability is used to reflect the degree to which the content entry passes the title display, image specifications, category mapping, and rule validation in the target channel.

[0056] For example, after acquiring the aforementioned features, the system first calculates multiple sub-scores separately, normalizes each sub-score to a unified dimension, and then merges them according to preset weighting rules to obtain a basic quality score. These preset weighting rules are not fixed but can be adjusted based on historical manual review results, post-release feedback, or error correction records, thus making the score more consistent with actual business needs. After completing the basic quality score, the system further identifies constraint conflicts and corrects the basic quality score accordingly. These constraint conflicts include at least missing key fields, size system conflicts, mismatches between digital assets and footwear master data, and strict disapproval by channel rules. In other words, if a content item scores highly in some dimensions but has issues such as empty key fields, obvious size mapping errors, discrepancies between the main image and the footwear description, or explicit prohibition by the target channel, the system will still deduct from the basic quality score to obtain a final quality score that better reflects the actual release risk.

[0057] For example, in this application, the system can first read the category, gender, size system, color, material, applicable scenarios, title and description of a certain shoe content item, and at the same time read its main product image, video cover and target channel configuration parameters; then, the system calculates the completeness score of shoe fields, the consistency score of size system, the consistency score of shoe attribute constraints, the matching score of digital assets and shoe master data and the channel adaptation score according to the aforementioned rules, and performs weighted fusion of each score to obtain a basic quality score. If the system detects that although the title, material, and color of an entry are complete, but the size data is inconsistent with the preset size mapping library, or the visual features extracted from the product's main image differ significantly from the shoe category represented by the title, then the corresponding situation will be identified as a constraint conflict, and the basic quality score will be corrected. Finally, the system will compare the corrected quality score with the preset publishing threshold and completion threshold. When the quality score reaches the publishing threshold, the content entry will be set to a publishable state; when the quality score is lower than the publishing threshold, the content entry will be set to a completion-required state, and the corresponding low-scoring dimension or conflict item will be recorded for subsequent completion and correction processing.

[0058] Through the above processing method, step four no longer judges the quality of content based on a single completeness indicator. Instead, it incorporates the footwear business attributes, the corresponding digital assets, and the target channel release requirements into a unified evaluation process. This allows the quality score results to more accurately reflect the actual publishability of the content items and provides a reliable basis for subsequent compliance review and release processes.

[0059] Step 5: Perform compliance review on content items that are ready for publication. The compliance review step uses a rule engine to ensure that the content items comply with the platform and legal requirements before being published to the target channel, thereby avoiding the spread of illegal information and content risks. When the review is passed, the content item is published. When the review fails, the system records the review evidence and saves the handling results for future reference.

[0060] Among them, the rule engine is a business rule execution framework that supports dynamic loading, hot updating, and interpretable output, used to perform multi-dimensional compliance verification; prohibited word matching refers to a sensitive word library scanning mechanism based on regular expressions or AC automata, used to identify sensitive or illegal words that may exist in the entry text; channel source blacklist matching refers to comparing the source identifier of the content entry with a preset restricted channel list, used to mark source compliance risks; grammatical structure legality judgment refers to performing DOM tree verification and sandbox parsing on the rich text fields in the content entry to detect illegal HTML tags, script elements, or event bindings and other illegal grammatical structures;

[0061] For example, in this application, the system configures the rule engine as a three-layer pipeline: the first layer performs a full-text scan of prohibited words on the content entries; if a violation is detected, publication is blocked, and the matching words and original text fragments are recorded; the second layer queries whether the entry's source identifier exists in the blacklist of high-risk channels; if so, it is marked as a source compliance risk; the third layer performs sandboxed parsing of HTML rich text fields, detecting script tags and other illegal syntax structures to ensure that the entries are grammatically legal, secure, and reliable when published. This configuration guarantees the compliance and traceability of content entries during multi-channel publication, while also supporting dynamic rule updates and expansions.

[0062] Step Six: Each time a content item is published, the system generates and stores a publication snapshot. This snapshot records and tracks the complete state of the published content, ensuring auditability, data consistency, and historical version management across multiple channels. The publication snapshot includes the channel publishing package, content item version number, operator identifier, sending time, and channel receipt. The snapshot is stored on a storage medium that meets a preset retention period to support subsequent rollback, traceability, and statistical analysis.

[0063] Among them, the channel release package refers to a content data package customized and packaged for a specific target channel to ensure that the format and content specifications of the entries are compatible across channels; the content entry version number is a string that follows the semantic versioning specification, which is used to identify the change level of the released entry relative to the original modeling version to support version control and change tracking; the operator identifier is used to record the system account or human account that performs the release operation for permission management and operation auditing; the channel receipt refers to the success response code, message ID or signature digest returned by the target channel to confirm whether the content has been successfully released.

[0064] For example, in this application, the system will store the published snapshots to a high-reliability database or object storage system, and can be configured with a redundant backup strategy to ensure that the complete publishing record can be quickly obtained in case of abnormal rollback or historical query; the snapshot data structure supports indexing according to content items, shoe style identifiers and publishing channels, thereby achieving efficient retrieval, statistics and analysis in a multi-channel environment, and further improving the intelligence and traceability of publishing management.

[0065] Example 2

[0066] See Figure 2 and Figure 3 The present invention also provides an internet content management system for the footwear industry, comprising: a data access module, a content modeling module, an alignment and deduplication module, a quality assessment and workflow module, a compliance review module, and a publishing module, which work together to complete the collection, organization, screening, review, and publishing management of footwear content;

[0067] The data access module is primarily responsible for importing data from multiple sources. This module receives content information from e-commerce platforms, brand-owned business systems, and social media platforms, and performs source tagging, time recording, and tenant differentiation during the integration process. For content from different sources, the module first performs basic formatting and field mapping, enabling data with different expression methods to enter a unified processing channel. For product attribute information, the module extracts its structured fields. For resources such as images, videos, or 3D models, it extracts their resource addresses, identification information, or associated descriptions for subsequent unified management.

[0068] After the integration is complete, the content modeling module merges and standardizes the input data. This module establishes unified content units around the shoe model object, organizing the attribute information representing the shoe model into the master data section, and organizing resources such as images, videos, and 3D models into the digital assets section, linking the two through a unified identifier. During processing, this module also standardizes field names, attribute value representations, enumeration items, and business meanings, ensuring that shoe data from different sources has a consistent data structure and business semantics after entering the system, providing a stable data foundation for subsequent deduplication and scoring.

[0069] After the standard content is formed, the alignment and deduplication module begins to identify and clean up potentially duplicate content. This module first groups the content based on shoe branding, then compares the similarity of attribute information within the same group. When branding is not entirely identical but content descriptions are highly similar, the module further considers attributes such as color, material, size system, and applicable scenarios for further judgment. Based on this, the alignment and deduplication module determines the final items to retain according to a preset retention strategy. For example, it prioritizes retaining data from more reliable sources, more recent updates, or manually verified data, while moving other duplicates to a backup status. For cases where complementary information exists across different sources, the module can also merge valid fields and resource references into the same content entry to reduce information loss.

[0070] After deduplication, the content enters the quality assessment and workflow module. This module includes a feature extraction unit for extracting footwear master data features, digital asset features, and channel release parameters; a sub-score calculation unit for calculating footwear field completeness scores, size system consistency scores, footwear attribute constraint consistency scores, digital asset-footwear master data matching scores, and channel adaptation scores; a fusion score unit for fusing the sub-scores according to preset weight rules to obtain a basic quality score; a score correction unit for correcting the basic quality score based on constraint conflict items to obtain a quality score; and a state switching unit for comparing the quality score with preset release and completion thresholds and triggering a workflow state change for the content item. This module first extracts the necessary assessment information from the content item, including footwear attributes, resource features, and the format requirements of the intended distribution channels. Subsequently, the module performs quality analysis on the item from multiple dimensions, including whether fields are complete, whether size settings are reasonable, whether attributes match each other, whether digital assets are consistent with the product master data, and whether the content meets the display specifications of the target channel. After completing the above analysis, the module synthesizes the results to form a basic score. If key items are missing, size logic is abnormal, the correspondence between images and text is invalid, or channel requirements are not met, further deductions and corrections are performed. The final results are used to drive the item into different processing states, such as directly entering the publishable process, or transferring it to subsequent paths such as completion and review. This module can also combine historical review status, error correction results, and publishing feedback to adjust the scoring strategy and judgment thresholds, making the scoring results closer to actual operational needs.

[0071] Once content enters the processing stage, the compliance review module performs rule verification. This module focuses on checking for risks in text expression, source security, and structural format. For text information such as titles, descriptions, and image / text descriptions, the module performs sensitive and restricted word checks; for source information, it checks for restricted sources or high-risk channels; for rich text or detail content, it verifies the legality of the structure to prevent abnormal tags, illegal scripts, or unsafe content from entering the publishing process. If necessary, the module also confirms the consistency and compliance of fields such as brand names and trademark information. Content that passes the review is sent to the publishing module, while content that fails the review retains corresponding evidence and processing records for subsequent investigation and correction.

[0072] Finally, the publishing module is responsible for outputting the approved content to the target channels. This module customizes and encapsulates the content items according to the display requirements of different channels, forming corresponding publishing data, and then initiates the sending to the target platform. After the content is sent, the publishing module synchronously generates a publishing record, saving key information from this output, including the published content, corresponding version, executor, sending time, and channel return results. This module stores the publishing record in a storage medium with retention capabilities, supporting subsequent queries by shoe style, version, channel, or operation time, thereby meeting business needs such as content traceability, publishing verification, anomaly analysis, and historical statistics.

[0073] Through the collaborative work of the above modules, the system in this embodiment can unify footwear content that was originally scattered across different sources, formats, and expressions into a single processing system, and sequentially complete standardization, deduplication, quality assessment, compliance review, and publication tracking, thereby improving the automation level, content quality, and multi-channel publishing efficiency of footwear industry content management.

[0074] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for internet content management in the footwear industry, characterized in that, Includes the following steps: S1. Obtain shoe industry-related content data from two heterogeneous data sources. The content data includes structured product data and unstructured digital asset data, and add a source identifier and timestamp to each piece of content data. S2. Based on a preset footwear content data model, the content data is parsed and standardized to generate content entries; wherein, the content entries include: a footwear master data domain and a digital asset domain. The footwear master data field contains footwear identifiers and a set of footwear attributes, and the digital asset field contains digital asset references associated with the footwear identifiers; S3. Based on the conflict resolution rules calculated by similarity, perform deduplication on the content entries; S4. Perform a quality assessment on the deduplicated content items to obtain a quality score, and trigger a workflow status change for the content items based on the quality score; S5. Perform compliance review on content items that are in the publishable or review-required state. The compliance review includes rule engine review. Publish the content item after the review is passed, and record the review evidence and handling results when the review fails. S6. Generate and store a release snapshot each time a release is made. The release snapshot includes the channel release package, content entry version number, operator identifier, sending time and channel receipt, and stores the release snapshot to a storage medium that meets the preset retention period.

2. The method for internet content management in the footwear industry according to claim 1, characterized in that, Step S4 includes: S41. Extract the footwear master data features, digital asset features, and channel release parameters of the content entries; S42. Based on the footwear master data characteristics, digital asset characteristics, and channel release parameters, calculate multiple sub-scores respectively. The sub-scores include at least the footwear field completeness score, size system consistency score, footwear attribute constraint consistency score, digital asset and footwear master data matching score, and channel adaptation score. S43. Merge the sub-scores according to the preset weighting rules to obtain the basic quality score; S44. Based on the constraint conflicts caused by missing key fields, size system conflicts, mismatch between digital assets and footwear master data, and hard non-compliance with channel rules, the basic quality score is corrected to obtain the quality score. S45. The quality score is compared with the preset publishing threshold and the completion threshold to trigger a change in the workflow status of the content item. The workflow status of the content item includes a publishable status and a completion-required status.

3. The method for internet content management in the footwear industry according to claim 1, characterized in that, The heterogeneous data sources include content from e-commerce platforms, brand-owned content sources, and social media content. The footwear attribute set includes category, gender, size system, color, material, and applicable scenarios.

4. The method for internet content management in the footwear industry according to claim 1, characterized in that, The conflict resolution rules include: retaining data based on source priority, retaining data based on the latest timestamp, retaining data based on manual confirmation marks, and merging and retaining content items from multiple sources based on shoe style identifiers.

5. A method for internet content management in the footwear industry according to claim 2, characterized in that, The completeness score for the footwear field was obtained in the following way: The importance coefficient of each field is determined according to the preset set of required fields, and the completeness score of the footwear field is calculated by the ratio of the sum of the importance coefficients of the filled and format-validated fields to the sum of the total importance coefficients of the preset set of required fields.

6. A method for internet content management in the footwear industry according to claim 2, characterized in that, The consistency score of the size system is obtained by matching the size data in the content items with the preset size mapping library; the consistency score of the footwear attribute constraints is obtained by verifying the preset constraint relationship between category, gender, size system and applicable scenario.

7. A method for internet content management in the footwear industry according to claim 2, characterized in that, The matching score between digital assets and footwear master data is obtained by extracting the visual or textual features of the digital assets and matching them with the title, description, or attribute tags in the footwear master data domain; the channel adaptation score is obtained by verifying the target channel's title length, image specifications, category mapping, and compliance with prohibited word rules.

8. A method for internet content management in the footwear industry according to claim 2, characterized in that, The preset weight rules, the release threshold, and the completion threshold are dynamically adjusted based on historical manual review results, post-release feedback results, or error correction records. The rule engine review process includes prohibited word matching, channel source blacklist matching, grammatical structure legality determination, and trademark or brand field compliance verification.

9. A method for internet content management in the footwear industry according to claim 1, characterized in that, The rule engine review process includes: matching prohibited words, matching the blacklist of channel sources, and determining the legality of the grammatical structure.

10. A content management system for the footwear industry internet platform, characterized in that, include: The data access module is used to obtain footwear industry-related content data from heterogeneous data sources and attach source identifiers, timestamps, and tenant identifiers. The content modeling module is used to generate content entries based on the footwear content data model. The alignment and deduplication module is used to deduplicate content items; The Quality Assessment and Workflow module is used to perform quality assessments on deduplicated content items, generate quality scores, and automatically set the workflow status of content items. The compliance review module is used to perform rule engine reviews on content items and record review evidence and processing results; The publishing module is used to publish content items, and at the same time generate a publishing snapshot and store it according to a preset retention period; The quality assessment and workflow module includes: The feature extraction unit is used to extract footwear master data features, digital asset features, and channel release parameters; Sub-scoring calculation units are used to calculate the completeness score of footwear fields, the consistency score of the size system, the consistency score of footwear attribute constraints, the matching score between digital assets and footwear master data, and the channel adaptation score, respectively. The fusion scoring unit is used to fuse the sub-scorings according to preset weighting rules to obtain a basic quality score; The scoring correction unit is used to correct the basic quality score based on the constraint conflict terms to obtain a quality score; The status switching unit is used to compare the quality score with preset publishing thresholds and completion thresholds, and trigger a workflow status change for the content item.

Images