Search

CN-121981798-A - AI monitoring evidence obtaining platform and equipment for discovering potential infringement behavior of online store

CN121981798ACN 121981798 ACN121981798 ACN 121981798ACN-121981798-A

Abstract

The application discloses an AI monitoring evidence obtaining platform and electronic equipment for discovering potential infringement behaviors of online stores, wherein the AI monitoring evidence obtaining platform is provided with a plurality of AI large models, after the AI monitoring evidence obtaining platform is started, text information of a prior right is input into a feature word to extract the large models to generate a feature word set, the feature word set is obtained through correction and supplementation of the feature word expansion large models by combining commodity naming, sales operation and transaction characteristics of known infringement online stores, the potential infringement commodity information is recalled on an e-commerce platform based on the expansion feature word set, right features of the prior right in a protection range are extracted through the infringement comparison large models and are compared with the potential infringement commodity information to obtain an infringement analysis result, if infringement is judged, evidence obtaining results are obtained for infringement commodities, and the evidence obtaining results are sent to a timestamp authentication service system to be added with a timestamp. The method and the system can improve the efficiency and accuracy of online store infringement evidence collection.

Inventors

  • ZHANG CHANGLI
  • Diao Chunfei

Assignees

  • 北京联合信任技术服务有限公司

Dates

Publication Date
20260505
Application Date
20251230

Claims (20)

  1. 1. The utility model provides an AI control evidence obtaining platform for discovering online store potential infringement action, its characterized in that, AI control evidence obtaining platform carries on a plurality of trained AI big models, AI control evidence obtaining platform and the timestamp authentication service system communication connection who presets, AI control evidence obtaining platform starts the back and carries out following processing: Inputting the text information of the previous right into a pre-trained feature word extraction large model, wherein the feature word extraction large model is used for generating a feature word set of the previous right according to the text information of the previous right; Inputting the generated feature word set into a pre-trained feature word expansion large model, wherein the feature word expansion large model is used for correcting or supplementing the feature word set by using naming information, sales operation and transaction feature information of one or more known infringement online stores on the known infringement commodities to obtain an expansion feature word set for searching the potential infringement commodities; Searching and recalling the potentially infringing commodity information in the electronic commerce platform based on the generated extended feature word set, wherein the potentially infringing commodity information comprises commodity description information and/or commodity pictures of the potentially infringing commodity; inputting the text information of the prior rights and each piece of potentially infringing commodity information into a pre-trained infringing comparison large model, wherein the infringing comparison large model is used for analyzing the text information of the prior rights to obtain a plurality of right features for defining the protection range of the prior rights; If the infringement analysis result is that the potentially infringing commodity is judged to be the infringing commodity, evidence obtaining is carried out on the infringing commodity, and the evidence obtaining result is sent to a time stamp authentication service system so that the time stamp authentication service system can add a time stamp to the evidence obtaining result.
  2. 2. The AI monitoring forensic platform according to claim 1 wherein the prior rights comprise issued patents, the text information of the prior rights comprises the name of the issued patent and the claims, and the issued patent comprises an utility model patent or an utility model patent; the feature word extraction large model is used to perform the following processing: Semantic analysis is performed on the name and the claim of the authorized patent, and product names, component features, connection relation features, material features and/or functional effect limiting features are extracted to be used as feature word sets of the authorized patent.
  3. 3. The AI monitoring forensic platform according to claim 2 in which the naming information of the known infringing article includes at least one of the name of the known infringing article, the name of the component in the known infringing article, and the name of the material in the known infringing article; The feature word expansion big model is specifically used for executing the following processes: if the name of the known infringed commodity is inconsistent with the name of the authorized patent, supplementing the name of the known infringed commodity serving as a parallel technical feature of the name of the authorized patent into an extended feature word set; And/or if the names of the components in the known infringing article are inconsistent with the names of the same components and/or the same materials in the authorized patent, taking the names of the components and/or the materials in the known infringing article as parallel technical characteristics of the names of the components and/or the materials in the authorized patent, and supplementing the parallel technical characteristics into the extended characteristic word set.
  4. 4. The AI monitoring forensic platform according to claim 2 or 3 wherein sales of the known infringing online store includes a description of the details of the goods for the known infringing goods, technical descriptions employed to circumvent patent infringement, functional effects and application scenarios in advertising literature and/or user consultation questions and answers; The feature word expansion big model is specifically used for executing the following processes: extracting descriptive words related to the functional effects to be protected by the authorized patent from sales call operation; Extracting from sales a substitute word describing the same application scenario as defined in the granted patent claims; Extracting ambiguous reference words from sales to components, structures or connections recited in an issued patent claim; the descriptive words, the alternative words and the ambiguous reference word are supplemented into the set of expanded feature words.
  5. 5. The AI monitoring evidence obtaining platform of claim 2 or 3, wherein the transaction characteristic information of the known infringing online store includes a selling price interval and/or a shipping place of the known infringing commodity sold by the known infringing online store; The feature word expansion large model is used to perform a process of supplementing a sales price interval and/or shipping location of a known infringing good into an expanded feature word set.
  6. 6. The AI monitoring forensic platform according to claim 1 wherein the feature word set contains only part of the rights features defining the scope of the preceding rights or the feature word set contains all of the rights features defining the scope of the preceding rights.
  7. 7. The AI-monitoring evidence-obtaining platform of claim 2, wherein the infringement contrast large model is specifically configured to: At least for the deconstructing of technical features from the independent claims of the issued patent, the independent claims are broken down into a plurality of minimum technical units, each minimum technical unit being a right feature comprising a component feature, a connection relationship feature, a material feature or a functional effect defining feature; Mapping and comparing the potentially infringing commodity information with the deconstructed multiple right features one by one; If the potentially infringing commodity information contains all the right features in the independent claims, the same infringing is judged to be formed, and an infringing analysis result that the potentially infringing commodity is an infringing commodity is obtained.
  8. 8. The AI monitoring forensic platform according to claim 7 wherein the infringement contrast large model is further for performing the following: If the potentially infringing article information does not contain all the right features of the independent claims, judging whether the potentially infringing article information contains equivalent features of the target right features for the target right features not contained in the potentially infringing article information; if the equivalent characteristics of the target right characteristics exist in the potentially infringing commodity information, the equivalent infringing is judged to be formed, and an infringing analysis result that the potentially infringing commodity is an infringing commodity is obtained.
  9. 9. The AI monitoring forensic platform according to claim 2 wherein the infringement analysis results further comprise at least one of infringement type, feature alignment table, confidence level, and infringement evidence guideline; The infringement type comprises the same infringement, equivalent infringement or non-infringement, the feature comparison table comprises similarity comparison details between a plurality of right features of the authorized patent and a plurality of corresponding features of the potentially infringement commodity, the confidence comprises certainty scores of infringement comparison big model to infringement type judgment conclusions, and the infringement evidence guide comprises text description or picture areas for feature comparison in information of the potentially infringement commodity.
  10. 10. The AI-monitoring forensic platform of claim 1 further configured, when performing forensics, to: And if the infringement analysis result is that the potentially infringing commodity is judged to be the infringement commodity, taking the infringement analysis result as a part of the evidence obtaining result.
  11. 11. The AI monitoring forensic platform according to claim 1 wherein the prior rights comprise a literal trademark and the textual information of the prior rights comprises the literal, registration number, approved registration category and goods or services specified for use of the literal trademark; the feature word extraction large model is used to perform the following processing: And carrying out semantic analysis on characters, registration numbers, approval registration categories and specified used goods or services of the characters and brands, and extracting standard forms, common variant forms, pinyin forms, foreign language translated name forms and general description words of industry-related goods or services of the characters and brands as feature word sets of the characters and brands.
  12. 12. The AI-monitoring evidence-obtaining platform of claim 11, wherein the naming information of the known infringing article includes an infringing article name resulting from adding, subtracting, homonym substitution, and/or form-near word substitution to the literal trademark; The feature word expansion big model is specifically used for executing the following processing: And if the name of the known infringement commodity is identified to contain the added word, the subtracted word, the homonym replacement or the shape-near word replacement of the literal trademark, taking the infringement commodity name as the parallel variant feature of the literal trademark, and supplementing the parallel variant feature into the expansion feature word set.
  13. 13. The AI-monitoring evidence-obtaining platform according to claim 11, characterized in that the sales of the known infringing online store is included in the advertisement document or the detailed description of the goods, the text content generated by inserting irrelevant symbols, letters, numbers or using their pinyin, foreign translation, abbreviation form or binding description with their known product model, advertisement language; The feature word expansion big model is specifically used for executing the following processing: extracting from sales art text variants containing the insertion of irrelevant symbols, letters or numbers to said literal trademark; extracting pinyin, foreign language translation or abbreviation form using the literal trademark from sales call operation; Extracting the keyword of the known product model or advertisement appearing in binding with the literal trademark from the sales operation; And supplementing the text variant, the pinyin, the foreign translation name or the abbreviated form and the known product model or the advertisement keyword into the extended feature word set.
  14. 14. The AI monitoring forensic platform according to claim 11 in which the infringement comparison large model is specifically for performing the following: Deconstructing characters, registration numbers, approval registration categories and specified used goods or services of the characters and trademarks to obtain a plurality of right features for defining the protection range of the characters and trademarks, wherein the right features comprise core font parts of the characters and trademarks, standard call pronunciation, recognized meanings and stable visual presentation patterns; Analyzing whether the potentially infringing commodity information contains a core font part of the literal trademark, and respectively calculating the similarity between the potentially infringing commodity information and standard call pronunciation, recognized meaning and stable visual presentation style of the literal trademark; And if the potentially infringing commodity information comprises a core font part of the literal trademark, and the similarity between the potentially infringing commodity information and the standard call pronunciation, the accepted meaning and the stable visual presentation style of the literal trademark is higher than a preset threshold, judging that the potentially infringing commodity is infringing, and obtaining an infringing analysis result of the potentially infringing commodity as an infringing commodity.
  15. 15. The AI monitoring forensic platform according to any one of claims 11 to 14 in which the infringement analysis results further comprise at least one of an infringement type, a feature alignment table, a confidence level, and an infringement evidence guideline; The infringement type comprises trademark infringement or trademark non-infringement, the feature comparison table comprises similarity comparison details between the potentially infringement commodity information and a core font part of a literal trademark, standard call pronunciation, a recognized meaning and a stable visual presentation style, the confidence comprises certainty scores of the infringement comparison large model on infringement type judgment conclusions, and the infringement evidence guide comprises text description or picture areas for feature comparison in the potentially infringement commodity information.
  16. 16. The AI monitoring forensic platform according to claim 1 wherein the prior rights comprise a copyrighted written work, the textual information of the prior rights comprising the author of the written work, the name of the work and at least a portion of the content piece of the work; The feature word extraction large model is used for executing the following processing: and carrying out semantic analysis on the author, the name of the work and at least part of the content fragments of the work, and extracting the author, the core originality expression fragments, the mark names appearing in the work, the characteristic plot structure keywords of the work and the literary style feature words of the author as a feature word set of the work.
  17. 17. The AI monitoring forensic platform according to claim 16 in which the naming information of the known infringing article includes the name of the known infringing article and the sales of the known infringing online store includes the article detail description and advertising literature of the known infringing article; The feature word expansion big model is specifically used for executing the following processes: Taking the name of the known infringed commodity, the detail description of the commodity or the synonymous rewritten segment, the word order adjusting segment, the expanded writing segment or the abbreviated segment of the core expression segment of the written work in the advertising document as the parallel characteristic of the core expression segment, and supplementing the parallel characteristic into the expanded characteristic word set; The name of the known infringement commodity, the detail description of the commodity or harmonic words, near-shape words, short names or others of the marked names of the written works in the advertising file are used as the parallel characteristics of the marked names and are supplemented into the expanded characteristic word set; And supplementing the related words of the names, commodity detail descriptions or advertisement cases of known infringement commodities, which are subjected to binding description with at least part of the content fragments, the names or the authors of the written works, into the extended feature word set.
  18. 18. The AI monitoring forensic platform according to claim 17 wherein the infringement contrast large model is specifically configured to perform the following: Performing idea and expression binary filtering on the names of the works and at least part of the content fragments of the works, and filtering out general facts, public domain information and idea concept parts which are not protected by copyright; performing semantic block segmentation and feature extraction on the residual original expression part after filtering to generate a plurality of right features, wherein each right feature corresponds to one original expression unit; giving weight coefficients determined based on originality height, core degree in the written works and creation difficulty to different right features; Respectively calculating text similarity between the potentially infringing commodity information and each right feature, and carrying out weighted calculation on the text similarity according to the weight coefficient to obtain a weighted similarity score; and if the weighted similarity score is greater than or equal to a preset weighted threshold, judging that the physical similarity is formed, and obtaining an infringement analysis result that the potential infringement commodity is an infringement commodity.
  19. 19. The AI monitoring forensic platform according to any one of claims 16 to 18 in which the infringement analysis results further comprise at least one of an infringement type, a feature alignment table, a confidence level, and an infringement evidence guideline; The infringement type comprises text similarity details of a plurality of right features of the written works and corresponding parts in the description information of the potentially infringing commodities, the confidence level comprises a certainty score of the infringement comparison big model to an infringement type judgment conclusion, and the infringement evidence guide comprises text paragraphs for feature comparison and corresponding positions of the text paragraphs in the written works, wherein the text paragraphs are pointed to the potentially infringement commodity information, and the text paragraphs are used for feature comparison.
  20. 20. The AI monitoring forensic platform according to claim 19 further for, in performing forensics, performing the following: Based on the text paragraphs pointed in the infringement evidence guide and the corresponding position information thereof, locating and extracting the corresponding text paragraphs in the infringement commodity and the written works, performing differential comparison and labeling, and generating a comparison report showing the same part, the substantially similar part and the differential part; Based on the differentiation comparison result, counting and generating quantized data, wherein the quantized data comprises the total number of the Lei-in characters, the number of Lei-in paragraphs and the proportion of Lei-in parts to corresponding chapters or the whole text of the written work; the comparison report is used with the quantized data as part of the forensic result.

Description

AI monitoring evidence obtaining platform and equipment for discovering potential infringement behavior of online store Technical Field The application relates to the technical field of infringement evidence obtaining, in particular to an AI monitoring evidence obtaining platform and electronic equipment for finding out potential infringement behaviors of online stores. Background With the vigorous development of the e-commerce industry, the number of online stores is increased drastically, and the problem of online store infringement is also increasingly highlighted. The traditional online store infringement evidence obtaining and processing mode has a plurality of defects. On the one hand, the traditional mode mainly relies on the manual work to carry out the investigation of online store one by one, commodity one by one, and not only inefficiency, omission appears moreover easily. Because of various goods and huge quantity on the electronic commerce platform, manual monitoring is difficult to comprehensively and timely examine massive goods information, so that many infringement behaviors cannot be found and restrained in time. On the other hand, traditional evidence-taking approaches lack effective time proof and authority. When infringement disputes occur, the validity and effectiveness of the evidence and the accuracy of the evidence obtaining time are important. The traditional evidence obtaining process is generally simple in screenshot, commodity page storing and the like, the evidence is easy to tamper, and the legally proving force is relatively weak. Moreover, due to the lack of a unified time authentication mechanism, it is difficult to accurately prove the time when the infringement occurs, which brings great difficulty to subsequent maintenance work. In addition, in order to evade supervision, an infringer often continuously changes the naming, description, sales, etc. of the commodities, so that it is difficult to accurately identify potential infringed commodities in a conventional search manner, such as based on fixed program codes (or fixed rules). For example, some infringers may use harmonic sounds, abbreviations, obscure expressions, etc. to replace the obvious infringement vocabulary, thereby bypassing traditional monitoring means. These hidden operations make it difficult for traditional online store infringement search evidence-obtaining methods to find infringed goods, increasing the difficulty and cost of evidence-obtaining. Disclosure of Invention In view of the above, the embodiment of the application provides an AI monitoring evidence obtaining platform and electronic equipment for discovering potential infringement behaviors of online stores, which are used for solving the technical problems of low online store infringement evidence obtaining efficiency and high difficulty. In a first aspect, the embodiment of the application provides an AI monitoring evidence obtaining platform for discovering online store potential infringement, the AI monitoring evidence obtaining platform is provided with a plurality of pre-trained AI big models, the AI monitoring evidence obtaining platform is in communication connection with a preset timestamp authentication service system, the AI monitoring evidence obtaining platform is started to execute the following processing that text information of a prior right is input into the pre-trained feature word extraction big model, the feature word extraction big model is used for generating a feature word set of the prior right according to the text information of the prior right, the generated feature word expansion big model is input into the pre-trained feature word expansion big model, the feature word expansion big model is used for correcting or supplementing the naming information, sales operation and transaction feature information of known infringement commodities by using one or more known infringement stores to obtain an expansion feature word set used for searching the potential infringement commodities, the potential infringement commodities are searched and retrieved in the electronic commerce platform based on the generated expansion feature word set, the potential infringement commodities comprise description information and/or commodity pictures of the prior right, the prior right and the pre-right feature word extraction big model is used for generating a feature word set of the prior right according to the text information of the prior right, the prior right and the large model is used for comparing the prior right and the security information with the security information to obtain a plurality of the protection result by comparing the prior right with the security information by the prior right and the item to obtain a comparison result, the comparison of the protection result is used for comparing the prior right and the protection information with the security information with the large item to obtain a comparison result, for the