WO2018120899A1 - 一种商标查询结果近似度评价和排序方法、装置 - Google Patents

一种商标查询结果近似度评价和排序方法、装置 Download PDF

Info

Publication number
WO2018120899A1
WO2018120899A1 PCT/CN2017/100187 CN2017100187W WO2018120899A1 WO 2018120899 A1 WO2018120899 A1 WO 2018120899A1 CN 2017100187 W CN2017100187 W CN 2017100187W WO 2018120899 A1 WO2018120899 A1 WO 2018120899A1
Authority
WO
WIPO (PCT)
Prior art keywords
trademark
card
divided
combination
indicates
Prior art date
Application number
PCT/CN2017/100187
Other languages
English (en)
French (fr)
Inventor
徐庆
Original Assignee
徐庆
佛山市国方商标服务有限公司
佛山市国方商标软件有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 徐庆, 佛山市国方商标服务有限公司, 佛山市国方商标软件有限公司 filed Critical 徐庆
Priority to US16/475,333 priority Critical patent/US20200387543A1/en
Publication of WO2018120899A1 publication Critical patent/WO2018120899A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/54Browsing; Visualisation therefor
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G06F16/5838Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using colour
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G06F16/5846Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using extracted text
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/903Querying
    • G06F16/90335Query processing
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/906Clustering; Classification
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/22Matching criteria, e.g. proximity measures
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00Handling natural language data
    • G06F40/20Natural language analysis
    • G06F40/237Lexical tools
    • G06F40/247Thesauruses; Synonyms
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00Handling natural language data
    • G06F40/20Natural language analysis
    • G06F40/279Recognition of textual entities
    • G06F40/284Lexical analysis, e.g. tokenisation or collocates
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00Handling natural language data
    • G06F40/20Natural language analysis
    • G06F40/279Recognition of textual entities
    • G06F40/289Phrasal analysis, e.g. finite state techniques or chunking
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/40Extraction of image or video features
    • G06V10/44Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/70Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/761Proximity, similarity or dissimilarity measures
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10Character recognition
    • G06V30/26Techniques for post-processing, e.g. correcting the recognition result
    • G06V30/262Techniques for post-processing, e.g. correcting the recognition result using context analysis, e.g. lexical, syntactic or semantic context
    • G06V30/274Syntactic or semantic context, e.g. balancing
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10Services
    • G06Q50/18Legal services
    • G06Q50/184Intellectual property management
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10Character recognition
    • G06V30/28Character recognition specially adapted to the type of the alphabet, e.g. Latin alphabet
    • G06V30/293Character recognition specially adapted to the type of the alphabet, e.g. Latin alphabet of characters other than Kanji, Hiragana or Katakana

Definitions

  • the invention relates to the field of trademark information retrieval, and particularly relates to a method and device for approximating and ranking the trademark query results.
  • trademark Chinese name characteristics trademark English name characteristics
  • pinyin letter features graphic element coding features
  • image feature descriptors etc.
  • the eigenvalues do not fully reflect the comprehensive characteristics of the combination of shape, sound and meaning of the trademark, resulting in the same or similar possible misjudgment of the trademark.
  • the result of the traditional trademark inquiry system is that the trademark sorting method is generally sorted according to a single feature, but the two or more features cannot be sorted in parallel. Therefore, the report and display sorting result trademarks have certain one-sidedness.
  • the Chinese invention patent with the application number 201410043915.0 is named: the trademark inquiry system and method, wherein the trademark inquiry system comprises: a query module for receiving the trademark to be inquired; and a feature extraction module for extracting the trademark feature of the trademark to be inquired An index library for storing the extracted trademark features of the trademark to be inquired; a trademark library for storing the existing trademark; a feature library for storing the trademark features of the existing trademark; and a retrieval module for The trademark feature of the query mark is matched with the trademark feature of the existing mark; and a display module is configured to display the result of the match.
  • the trademark inquiry system comprises: a query module for receiving the trademark to be inquired; and a feature extraction module for extracting the trademark feature of the trademark to be inquired
  • An index library for storing the extracted trademark features of the trademark to be inquired
  • a trademark library for storing the existing trademark
  • a feature library for storing the trademark features of the existing trademark
  • a retrieval module for The trademark feature of the query mark is matched with
  • Paragraph 0043 of the patent specification discloses a method for calculating the degree of approximation of the existing trademark or a method for realizing the same: the retrieval module 106 is mainly used to implement the retrieval matching process, and according to the correlation calculation method, the matching and screening of the trademark is realized, and finally, the matching requirement is obtained. The results are fed back to the user.
  • the retrieval module 106 provides a retrieval interface to the user based on the query of the content, translating the retrieval request of the user into a question that can operate on the database. Searching allows for global objects, such as the entire trademark, as well as for sub-objects in it, as well as any combination.
  • the results returned by the retrieval module 106 can be outputted in accordance with the degree of similarity, and the display module 107 can display the sorted existing trademarks and, if necessary, perform further queries based on the obtained retrieval results. Since content-based retrieval implements similarity retrieval, mimicking human cognitive processes, it is also necessary to refine the retrieval results in constant interaction with the retrieval user.
  • the above patented technical solution can only solve the problem that the matching degree of matching of the user's single or one-by-one retrieval request is separately sorted, but cannot solve the problem of comprehensive ranking which can be matched by multiple retrieval requests and can match the degree of similarity of multiple retrieval requests.
  • Some single features of some trademarks cannot fully reflect the comprehensive characteristics of the combination of shape, sound and meaning of trademarks.
  • the results of similarity ranking by single features may not meet the requirements of the same or similar trademarks in the sense of Trademark Law.
  • the similarity ranking results may cause users of the trademark inquiry system to mistakenly believe that the pre-existing trademark may be the same or similar to the trademark in the sense of the Trademark Law, which may lead to serious mistakes in trademark registration, management and protection.
  • the existing trademark inquiring method the ordering of the trademark approximation needs to continuously interact with the users of the trademark inquiring system to provide a sorting result of a plurality of different features matching similarity for the user's reference, and also increases the user.
  • the workload of the query operation the ordering of the trademark approximation needs to continuously interact with the users of the trademark inquiring system to provide a sorting result of a plurality of different features matching similarity for the user's reference, and also increases the user.
  • the workload of the query operation is the ordering of the trademark approximation needs to continuously interact with the users of the trademark inquiring system to provide a sorting result of a plurality of different features matching similarity for the user's reference.
  • the object of the present invention is to provide a method and apparatus for approximating and ranking the trademark query result approximation, which can obtain a comprehensive quantitative value of the trademark approximation for comprehensively evaluating the searched result trademark and the input trademark in multiple features. And according to the size of the comprehensive quantitative value, the result trademarks are sorted, so that the result trademarks seen by the users are more in line with the same or similar requirements of the trademarks in the sense of the "Trademark Law", and avoiding the single feature sorting cannot fully reflect the various characteristics of the trademarks. Defects such as misreporting and misreporting of trademark searches.
  • a method for approximating and ranking the trademark query results, and approximating and sorting the approximate trademark query results including the following steps:
  • Step S110 Performing a trademark card processing on the sample trademark image and content according to a preset trademark card division standard, and the specific processing process includes: (1) establishing a plurality of combinations of minimum shape elements of preset shape features, pronunciation features, and meaning features.
  • the trademark sub-card standard constituted by the scheme, (2) whether the sample trademark is identified by Chinese characters, graphics, letters, numbers or symbols, and the content of the constituent elements is obtained, (3) the shape of each component of the sample trademark The minimum feature unit, the minimum unit of the pronunciation feature and the minimum unit of the meaning feature; (4) extracting the segmentation information of various characters and graphics generated or converted by each combination scheme according to the established trademark card classification standard, and The segmentation information is used as the sample trademark card information, and the approximation evaluation score of each preset trademark card standard is set;
  • Step S120 Performing a trademark card processing on the input trademark image and content according to a preset trademark card standard, and the specific processing process includes: (1) establishing a plurality of combinations of minimum shape elements, preset sound features, and meaning features.
  • the trademark sub-card standard constituted by the scheme, (2), whether the input trademark is identified by Chinese characters, graphics, letters, numbers or symbols, and the content of the constituent elements is obtained; (3) extracting the constituent elements of the input trademark Shape feature minimum unit, pronunciation feature minimum unit and meaning feature minimum unit; (4), according to the established trademark card standard, extract the segmentation information of various characters and graphics generated or converted by each combination scheme, These segmentation information is used as input trademark card information;
  • Step S130 Searching the sample trademark card information stored in the trademark storage by using the input trademark card information set as a search keyword, and acquiring the card information and the card matching information of the relevant result trademark;
  • Step S140 Calculating a formula according to a preset trademark near rate, a trademark proximity rate, a trademark sound near rate, and a search keyword matching score rate, respectively calculating a trademark proximity rate and a trademark proximity rate between the input trademark and the result trademark. , trademark tone near rate and search keyword matching score rate;
  • Step S150 Calculate the comprehensive quantized value of the trademark approximate degree according to the calculation formula of the preset trademark approximate degree comprehensive quantitative value, and then sort the result trademark by using the size of the trademark approximate degree comprehensive quantized value.
  • the smallest unit of shape features includes:
  • the constituent elements are the smallest unit of shape features of Chinese characters, and one of them is selected as follows: each Chinese character, each stroke of each Chinese character;
  • the constituent element is the smallest unit of the shape feature of the graphic, and one of the following is selected: the trademark graphic element code, and the set of pixel points on the outline of the long trademark image is preset;
  • the constituent element is the smallest unit of the shape feature of the letter, and one of the following is selected: a word for each language, and each letter;
  • the constituent elements are the smallest unit of shape features of Chinese numerals, and one of them is selected as follows: a combination of Chinese numbers, each individual Chinese number;
  • the smallest element of the shape feature of the Arabic numerals is selected as follows: a combination of Arabic numerals, each individual Arabic number;
  • the smallest element of the shape feature of the other language numbers is selected as follows: a combination of other language numbers, each individual other language number;
  • the constituent elements are the smallest feature of the shape feature of the symbol: for each individual symbol.
  • the constituent elements are the smallest unit of meaning of Chinese characters: the overall Chinese character combination of the trademark is recorded in the Chinese dictionary.
  • each word is the smallest unit of meaning feature; otherwise, the overall Chinese character combination of the trademark is the smallest unit of meaning feature;
  • the constituent element is the meaning of the graphic.
  • the constituent elements are the meaning of the letters.
  • the overall letter combination is the smallest unit of meaning features;
  • the constituent elements are the smallest unit of meaning meaning of Chinese numerals. One of them is selected as follows: the default reference language number corresponding to each group of Chinese characters separated in the trademark, and the default reference language number corresponding to each single Chinese digit in the trademark. Wherein the predetermined reference language number is any language number;
  • the constituent element is the smallest unit of meaning of Arabic numerals.
  • One of the following is selected: the default reference language number corresponding to each group of Arabic numerals separated in the trademark, and the default reference language number corresponding to each individual Arabic number in the trademark.
  • the predetermined reference language number is any language number;
  • each set of other language numbers separated in the trademark the corresponding preset reference language number, corresponding to each individual language number in the trademark
  • Presetting a reference language number Presetting a reference language number, the preset reference language number being any language number
  • the constituent elements are the meaning of the symbol.
  • the minimum unit of pronunciation features includes:
  • the constituent elements are the smallest unit of the pronunciation features of Chinese characters, which are the pinyin of each Chinese character;
  • the constituent element is the smallest unit of the pronunciation feature of the graphic, and is the pinyin of the name of each thing corresponding to the coding of the trademark graphic element;
  • the constituent element is the smallest unit of the pronunciation feature of the letter, and one of the following is selected: the pronunciation of each letter combination, the pronunciation of each letter;
  • the smallest unit of the pronunciation feature whose constituent elements are numbers or symbols, one of which is selected: the pronunciation of each group of numbers separated in the trademark, the pronunciation of each individual number, the pronunciation of each group of symbols separated in the trademark, each The pronunciation of a single symbol.
  • the trademark card standard includes:
  • the trademark card standard consisting of multiple combinations of shape features and minimum elements of Chinese characters, including: card standard a 1 , a 2 , a 3 , a 4 , a 5 , a 6 , a 7 At least one of a 8 , a 9 , a 10 , a 11 , a 12 , a 13 , wherein
  • a 1 indicates that the combination of all the language characters and graphic element codes of the entire trademark is divided into one sub-card.
  • a 2 indicates that the combination of all the language characters and graphic element codes of the trademark as a whole in reverse order is divided into one card.
  • a 3 means that the Chinese characters in the trademarks arranged in order are divided into one card
  • a 4 indicates that the Chinese characters in the trademarks arranged in reverse order are divided into one sub-card
  • a 5 means that the Chinese characters in the trademarks are divided into one sub-cards.
  • a 6 indicates that the Chinese alphabet in the reverse order is divided into a sub-card.
  • a 7 means that each relatively independent part of the trademark is divided into one sub-card
  • a 8 means that the trademark text contains the existing Chinese word mark completely, and the part is divided into one card.
  • a 9 means that the traditional and foreign characters contained in the trademark are converted into simplified characters and then divided into one card.
  • a 10 means that each word in the trademark is replaced by a sub-card after being replaced by a near-word.
  • a 11 indicates that each adjacent Chinese character in the trademark is divided into one sub-card
  • a 12 indicates that the first Chinese character combination in the trademark is divided into one sub-card
  • a 13 means that each Chinese character in the trademark is divided into one sub-card
  • a trademark card standard consisting of a plurality of combinations of shape features and minimum elements of letters, numbers, and symbols, including: card standard b 1 , b 2 , b 3 , b 4 , b 5 , b At least one of 6 , b 7 , b 8 , b 9 , b 10 , b 11 , b 12 , b 13 , b 14 , wherein
  • b 1 indicates that the combination of all the language characters and graphic element codes of the trademarks arranged in order is divided into one sub-card.
  • b 2 indicates that the combination of all the language characters and graphic element codes of the trademark as a whole in reverse order is divided into one card.
  • b 4 indicates that the letter combination in the reverse order of the trademark is divided into a sub-card
  • b 5 indicates that the non-Chinese numbers contained in the listed trademarks or each individual non-Chinese number are respectively divided into one sub-cards.
  • b 6 indicates that the non-Chinese numbers contained in the reverse-ordered trademarks or each individual non-Chinese number are respectively divided into one sub-cards.
  • b 10 means that each letter in the trademark is replaced by a close letter and then divided into a minute card.
  • b 12 indicates that the letters in the trademark are arranged in different order and then divided into one sub-cards.
  • b 13 indicates that the first and last letter combination in the trademark is divided into one sub-card.
  • a trademark card standard composed of a plurality of combination schemes of a minimum feature of a shape feature of a graphic, comprising: at least one of the card standard c 1 , c 2 , c 3 , and c 4 , wherein
  • c 1 indicates that the trademark graphic element code set is divided into one sub-card as a whole.
  • c 3 denotes that the trademark image feature descriptor generated by each image feature recognition method is respectively divided into one card
  • c 4 indicates that the pre-set length of the trademark image feature descriptor generated by each image feature recognition method is respectively divided into one sub-card, and the pre-set length of the trademark image feature descriptor refers to a pre-set trademark image outline.
  • the length of consecutively connected pixels, consecutively connected pixels are represented by a feature string set or a digital set, and the value ranges from 0.1% to 50% of the trademark image feature descriptor or the total length of the digital set;
  • a trademark card standard composed of a plurality of combination schemes of minimum components of the pronunciation features of Chinese characters, including: at least one of the card standard d 1 , d 2 , and d 3 , wherein
  • d 1 indicates that the Pinyin syllable of each Chinese character in the trademark is divided into one sub-card.
  • d 2 indicates that the pinyin corresponding to the overall Chinese character in the trademark is divided into a sub-card.
  • d 3 indicates that each Chinese character in the trademark is replaced with a near-word and the pinyin syllable is divided into one sub-card;
  • a trademark card standard composed of a plurality of combination schemes of a minimum unit of pronunciation features of letters, numbers, and symbols, including: at least one of the card standard e 1 , e 2 , e 3 , and e 4 , among them,
  • e 1 means that the pronunciation syllable of each English word in the trademark is divided into a sub-card.
  • e 2 means that the whole letter combination obtained by replacing the letter combination in the trademark by the close letter combination is divided into one card
  • e 3 indicates that the pronunciation syllable of each digit in the trademark is divided into a sub-card.
  • e 4 indicates that the pronunciation syllable of each symbol in the trademark is divided into one sub-card
  • a trademark card standard composed of a plurality of combination schemes of minimum components of a pronunciation feature composed of a constituent element, including: a card division standard f 1 , wherein f 1 represents the name of each thing corresponding to the coding of the graphic element of the trademark The pinyin is divided into a sub-card;
  • a trademark card standard composed of a plurality of combinations of meaning elements having a minimum feature unit of Chinese characters, including: at least one of the card standard g 1 , g 2 , g 3 , and g 4 , wherein
  • g 1 means that the trademark contains the existing Chinese word mark in the trademark server, and the whole trademark has no meaning.
  • the part containing the existing Chinese word mark is divided into one card.
  • g 2 indicates that the vocabulary included in the Chinese dictionary and the Chinese character combination of the existing Chinese character trademark in the trademark server all match, and the matching part is respectively divided into one sub-card.
  • g 3 indicates that the Chinese vocabulary contained in the trademark is replaced by a synonym and is divided into a sub-card.
  • g 4 indicates that the trademark as a whole has no meaning, and the overall Chinese character is divided into one sub-card;
  • H A trademark card standard consisting of a plurality of combinations of meaning components of a combination of letters, numbers and symbols, including sub-card standards h 1 , h 2 , h 3 , h 4 , h 5 , h 6 At least one of h 7 , h 8 , h 9 , wherein
  • h 1 indicates that the overall letter combination of the trademark is composed of a combination of words recorded in an English dictionary or other language dictionary, and the overall word combination is divided into one sub-card.
  • h 2 indicates that the trademark contains words in the English dictionary or other language dictionary, and each word is divided into one card.
  • h 3 indicates that the trademark contains words in the English dictionary or other language dictionary, and the synonyms of each word are divided into one card.
  • h 4 indicates that the overall letter combination of the trademark does not match the words recorded in the English dictionary or other language dictionary, and the overall letter combination is divided into one sub-card.
  • h 5 means that each group of numbers separated in the trademark is divided into one card.
  • h 6 indicates that the overall digital combination of the trademark is divided into one sub-card
  • h 7 indicates that the overall symbol combination of the trademark is divided into one sub-card
  • h 8 means that each symbol of the trademark is divided into one sub-card
  • h 9 means that the trademark completely contains the existing letter combination trademark of the trademark server, and the whole trademark has no meaning, and the part containing the existing letter combination trademark is divided into one sub-card;
  • the constituent elements are the meaning of the graphic features.
  • the minimum unit is a trademark card standard composed of a plurality of combination schemes, and includes at least one of the card standard i 1 and i 2 , wherein
  • i 1 indicates that the name of each thing corresponding to the trademark graphic element code is divided into one card
  • i 2 indicates that the trademark image feature descriptor corresponds to the trademark graphic element code, and each transaction name corresponding to the trademark graphic element code is divided into one sub-card.
  • the component is a minimum unit of the exception adjustment text
  • the trademark card division standard comprises at least one of the card standard y 1 and y 2 , wherein
  • y 1 means that the trademark contains the exception adjustment text, and the exception adjustment text is divided into one sub-card as a whole;
  • y 2 means that the trademark contains the exception adjustment text, and the exception adjustment text is divided into one sub-card.
  • the exception adjustment text includes one or more presets as follows: a place name of an administrative area above the county level, a foreign place name known to the public, and a general product name, indicating the quality of the product, the main raw materials, functions, uses, weights, quantities, and Other characteristics of the word, the generic name of the goods service, the text of the weak character.
  • the significantly weaker text refers to some custom texts that do not have the distinctive features of the trademark.
  • the exception adjustment text is recorded in the base name dictionary library, including: a world country and region dictionary table, a county-level administrative region name dictionary table, a foreign city name dictionary table, a banned word dictionary table, and the like.
  • the trademark query result approximation evaluation and sorting method wherein the “input trademark card information” in the step S120 includes: U 0 , ⁇ 1 , V 0 , ⁇ 2 , M 0 , Y 0 , wherein U 0 represents Enter the number of cards obtained based on the trademark card standard a 13 , b 14 , c 2 , c 4 or a combination thereof; ⁇ 1 represents the card based standard a 13 , b 14 , c of the exception adjustment text contained in the input trademark 2 , c 4 score card number or number of characters; V 0 represents the number of cards obtained by the input trademark based on the trademark card standard d 1 , d 2 , d 3 , e 1 , e 2 , e 3 , e 4 or a combination thereof; ⁇ 2 represents the number of cards or syllables based on the card-based standards d 1 , d 2 , d 3 , e 1 , e 2
  • the sub-card information and the sub-card matching information of the result trademark described in step S130 include Y a , U a , U b , U c , V a , V b , V c , M 1 , M 2 , M 3 , M 4 .
  • Y a represents the number of cards of the resulting trademark based on the trademark card standard y 1 or y 2 ;
  • U a represents the result of the trademark removal exception adjustment text and the input trademark The number of matching cards in the score card based on the trademark card standard a 13 , b 14 , c 2 , c 4 or a combination thereof;
  • U b indicates the result of the trademark removal exception adjustment text and the input trademark based on the trademark card standard a 10 , b 10 or a combination thereof, the number of matching cards in the score card;
  • U c indicates that the resulting trademark and the input trademark are based on the trademark card standard a 13 , b 14 , c 2 , c 4 or a combination thereof and a 10 , b 10 Or the combination of the scorecards obtained by the combination thereof is inserted into the number of mismatched cards;
  • V a represents the result of the trademark removal exception adjustment text and the input trademark is
  • M 3 indicates the result of the comparison after the trademark removal exception adjustment text and the input trademark Based on the number of matching cards of the trademark card standard g 3 , M 4 indicates the result of the comparison.
  • the type of the card obtained by classifying the trademark card information by the preset classification standard.
  • the feature type is divided according to the shape and meaning of the shape, including: a shape feature type, a pronunciation feature type, and a meaning feature type; and the content division according to the constituent elements includes: a Chinese character feature type, an alphabet character feature type, a digital text feature type, a symbol text feature type, Graphic element coding graphic feature type, image feature descriptor graphic feature type.
  • the trademark query result approximation evaluation and sorting method wherein the preset formula for calculating the trademark proximity rate, the trademark proximity rate, the trademark sound near rate, and the search keyword matching score rate in the step S140 includes:
  • W unit U a /(U 0 - ⁇ 1 )+[U b /(U 0 - ⁇ 1 )] ⁇ 1 -[U c /(U 0 - ⁇ 1 )] ⁇ 2 ,
  • W unit represents a trademark near rate
  • ⁇ 1 and ⁇ 2 are preset adjustment weights
  • values of ⁇ 1 and ⁇ 2 are between 10% and 300%;
  • S sound represents the trademark sound closeness
  • ⁇ 1 and ⁇ 2 are preset adjustment weights, and the value ranges are between 10% and 300%;
  • the formula for calculating the trademark near rate includes:
  • ⁇ 1 , ⁇ 2 , ⁇ 3 respectively represent the adjustment parameters for M 2 , M 3 , M 4 , and the value rule: when M 1 , M 2 , M 3 , M 4 appear
  • the first parameter in M 1 , M 2 , M 3 , M 4 is the effective parameter, and the rest is the invalid parameter.
  • ⁇ 1 , ⁇ 2 , ⁇ 3 has a value of 0; when M 1 is 0 and M 2 is not 0, ⁇ 1 is 1, ⁇ 2 , ⁇ 3 have a value of 0; when M 1 , M 2 are 0 and M 3 is not 0 When ⁇ 2 is 1, ⁇ 3 is 0; when M 1 , M 2 , M 3 is 0 and M 4 is not 0, ⁇ 3 is 1; ⁇ indicates that the input trademark is different from the comparison result. Parameter, ranging from 1% to 90%;
  • the search keyword matching score rate calculation formula includes at least one of the following: the search keyword matching comprehensive average score rate, the search keyword matching category average score rate, the search keyword matching category highest score rate, and the search keyword matching category weighting highest. Scoring rate, ie:
  • S keywork represents the search keyword matching score rate
  • S 1 represents the search keyword matching comprehensive average score rate
  • S 2 represents the search keyword matching category average score rate
  • S 3 represents the search keyword matching category highest score rate
  • S4 represents Search keyword matching classification weighted highest score rate
  • ⁇ 1 , ⁇ 2 , ⁇ 3 , ... ⁇ i respectively represent the scores of the result trademark and the input trademark in the first feature type, the second feature type, the third feature type, ..., the i-th feature type.
  • the calculation of the highest score in the preset approximation evaluation score of the card corresponding to the card, ⁇ 1 , ⁇ 2 , ⁇ 3 , ... ⁇ i ranges from 1% to 80% The total of all calculated weights is 100%.
  • the trademark query result approximation evaluation and sorting method wherein the formula for calculating the trademark approximate degree comprehensive quantized value in step S150 comprises:
  • TM near W unit ⁇ Q 1 +S sound ⁇ Q 2 +S meaning ⁇ Q 3 +S keywork ⁇ Q 4 ,
  • TM near represents the comprehensive quantitative value of trademark approximation
  • W unit represents the trademark proximity
  • S sound represents the trademark sound near rate
  • S meaning represents the trademark proximity rate
  • S keywork represents the search keyword matching score rate
  • Q 1 , Q 2 , Q 3 and Q 4 respectively represent the weight of the trademark near rate
  • the Q 1 , Q 2 , Q 3 and Q 4 values range from 5 Between % and 95%, the total of all calculated weights is 100%.
  • the present invention also provides an apparatus for approximating and sorting processing of trademark query results, comprising:
  • Sample trademark card pre-processing module used to process the trademark card image and content according to the preset trademark card standard.
  • the specific processing process includes: (1) establishing a preset shape feature, pronunciation features and Meaning-characteristics The minimum number of units of the trademark division card standard, (2), whether the sample trademark is identified by Chinese characters, graphics, letters, numbers or symbols, to obtain the content of the constituent elements, (3), The minimum feature unit of the shape characteristic of each component of the sample trademark, the smallest unit of the pronunciation feature and the smallest unit of the meaning feature; (4) extracting various characters and graphics generated or converted by each combination scheme according to the established trademark card classification standard The segmentation information, the segmentation information is used as the sample trademark card information, and the approximation evaluation score of each preset trademark card standard is set;
  • the input trademark card processing module is used for processing the trademark image and the content according to the preset trademark card standard.
  • the specific processing process includes: (1) establishing a preset shape feature, pronunciation feature and meaning (2) The identification of whether the input trademark is composed of Chinese characters, graphics, letters, numbers or symbols, and the content of the constituent elements; (3), extraction Entering the minimum feature of the shape feature of each component of the trademark, the minimum unit of the pronunciation feature and the minimum unit of the meaning feature; (4) extracting various characters and graphics generated or converted by each combination scheme according to the established trademark card classification standard The segmentation information, using the segmentation information as the input trademark card information;
  • the trademark search module searches for the sample trademark card information stored in the trademark storage by using the input trademark card information set as a search key, and obtains the card information and the card matching information of the relevant result trademark;
  • the trademark shape near rate calculation module is configured to calculate a trademark form close ratio between the input trademark and the result trademark according to a preset formula for calculating the trademark shape near rate;
  • the trademark near-rate calculation module is used to calculate the trademark proximity ratio between the input trademark and the result trademark according to the preset formula of the trademark near-probability ratio;
  • Trademark sound near rate calculation module used to calculate the trademark sound near rate between the input trademark and the result trademark according to the preset trademark sound near rate calculation formula
  • the search keyword matching score rate calculation module is configured to calculate a search keyword matching score rate between the input trademark and the result trademark according to a preset search keyword matching score rate calculation formula
  • the calculation module of the comprehensive quantized value of the trademark approximation used to calculate the comprehensive quantitative value of the trademark approximation according to the preset formula of the comprehensive approximation of the trademark approximation, and then use the approximate value of the trademark approximation to the result trademark Row sorting.
  • the present invention uses the preset trademark card classification standard to separately segment the input trademark from different angles to obtain the smallest unit and combination of shape features, pronunciation features and meaning features, and calculate the retrieval key between the result trademark and the input trademark.
  • Word matching score rate, shape near rate, sound near rate and righteousness ratio obtain the comprehensive quantitative value of trademark approximation and sort according to the approximate degree of comprehensive quantized value, which can fully reflect the comprehensive feature approximation of the shape, sound and meaning of trademark.
  • the invention only needs to input the trademark to be retrieved into the system once to obtain the best comprehensive sorting result, and overcomes the need for the existing trademark retrieval system to continuously perform human-computer interaction to obtain different sorting and display results, or use artificial The results of the screening result are too subjective.
  • Fig. 1 is a flow chart showing the method for approximating and ranking the trademark query results according to the first embodiment of the present invention.
  • Fig. 2 is an exemplary original view of the trademark of the first embodiment of the present invention.
  • FIG. 3 is an image feature descriptor diagram of the pixel of the trademark image on the outline of the image of the image of FIG. 2n using the 10 ⁇ 10 coordinate system standard.
  • FIG. 4 is an image feature descriptor diagram of the pixel of the trademark image on the outline of the image of the image of FIG. 2n using the 20 ⁇ 20 coordinate system standard.
  • FIG. 5 is a screenshot of the first 24 results trademark report interface sorted by the trademark approximate degree comprehensive quantization value in the embodiment 1 of the present invention.
  • Fig. 6 is a block diagram showing the structure of the approximation evaluation and sorting apparatus for trademark search results according to the first embodiment of the present invention.
  • Fig. 7 is a flow chart showing the method for approximating and ranking the trademark query results according to the second embodiment of the present invention.
  • a trademark query result approximation evaluation and ranking method includes the following steps:
  • Step S110 Performing a trademark card processing on the sample trademark image and content according to a preset trademark card division standard, and the specific processing process includes: (1) establishing a plurality of combinations of minimum shape elements of preset shape features, pronunciation features, and meaning features.
  • the trademark sub-card standard constituted by the scheme, (2) whether the sample trademark is identified by Chinese characters, graphics, letters, numbers or symbols, and the content of the constituent elements is obtained, (3) the shape of each component of the sample trademark The minimum feature unit, the minimum unit of the pronunciation feature and the minimum unit of the meaning feature; (4) extracting the segmentation information of various characters and graphics generated or converted by each combination scheme according to the established trademark card classification standard, and The segmentation information is used as the sample trademark card information, and the approximation evaluation score of each preset trademark card standard is set;
  • Step S120 Performing a trademark card processing on the input trademark image and content according to a preset trademark card standard, and the specific processing process includes: (1) establishing a plurality of combinations of minimum shape elements, preset sound features, and meaning features.
  • the trademark sub-card standard constituted by the scheme, (2), whether the input trademark is identified by Chinese characters, graphics, letters, numbers or symbols, and the content of the constituent elements is obtained; (3) extracting the constituent elements of the input trademark Shape feature minimum unit, pronunciation feature minimum unit and meaning feature minimum unit; (4), according to the established trademark card standard, extract the segmentation information of various characters and graphics generated or converted by each combination scheme, These segmentation information is used as input trademark card information;
  • Step S130 Searching the sample trademark card information stored in the trademark storage by using the input trademark card information set as a search keyword, and acquiring the card information and the card matching information of the relevant result trademark;
  • Step S140 Calculating a formula according to a preset trademark near rate, a trademark proximity rate, a trademark sound near rate, and a search keyword matching score rate, respectively calculating a trademark proximity rate and a trademark proximity rate between the input trademark and the result trademark. , trademark tone near rate and search keyword matching score rate;
  • Step S150 Calculate the comprehensive quantized value of the trademark approximate degree according to the calculation formula of the preset trademark approximate degree comprehensive quantitative value, and then sort the result trademark by using the size of the trademark approximate degree comprehensive quantized value.
  • step S110 the sample trademark image and the content are processed according to the preset trademark card classification standard, and the specific processing process includes: (1) establishing a minimum unit of the preset shape feature, the pronunciation feature and the meaning feature. (2) Identify whether the sample trademark is composed of Chinese characters, graphics, letters, numbers or symbols, and obtain the content of the constituent elements. (3) Composition of sample trademarks The minimum feature unit of the shape feature, the smallest unit of the pronunciation feature and the minimum unit of the meaning feature; (4) extracting the segmentation information of various characters and graphics generated or converted by each combination scheme according to the established trademark card classification standard The cut information is used as sample trademark card information, and the approximation evaluation score of each preset trademark card standard is set.
  • the embodiment of the present invention establishes a trademark card standard by subdividing the smallest constituent unit of the trademark in terms of shape, meaning, and pronunciation, and can obtain beneficial technical effects in the approximate evaluation and sorting process of the trademark query result. .
  • the smallest unit of shape features includes:
  • the constituent elements are the smallest unit of shape features of Chinese characters, and one of them is selected as follows: each Chinese character, each stroke of each Chinese character;
  • the constituent element is the smallest unit of the shape feature of the graphic, and one of the following is selected: the trademark graphic element code, and the set of pixel points on the outline of the long trademark image is preset;
  • the constituent element is the smallest unit of the shape feature of the letter, and one of the following is selected: a word for each language, and each letter;
  • the constituent elements are the smallest unit of shape features of Chinese numerals, and one of them is selected as follows: a combination of Chinese numbers, each individual Chinese number;
  • the smallest element of the shape feature of the Arabic numerals is selected as follows: a combination of Arabic numerals, each individual Arabic number;
  • the smallest element of the shape feature of the other language numbers is selected as follows: a combination of other language numbers, each individual other language number;
  • the constituent elements are the smallest feature of the shape feature of the symbol: for each individual symbol.
  • the constituent elements are the smallest unit of meaning of Chinese characters: when the overall Chinese character combination of a trademark is composed of a combination of words recorded in a Chinese dictionary, each word is the smallest unit of meaning features. Otherwise, the overall Chinese character combination of the trademark is the smallest unit of meaning features. ;
  • the constituent element is the meaning of the graphic.
  • the constituent elements are the meaning of the letters.
  • the overall letter combination is the smallest unit of meaning features;
  • the constituent elements are the smallest unit of meaning meaning of Chinese numerals. One of them is selected as follows: the default reference language number corresponding to each group of Chinese characters separated in the trademark, and the default reference language number corresponding to each single Chinese digit in the trademark. Wherein the predetermined reference language number is any language number;
  • the constituent element is the smallest unit of meaning of Arabic numerals.
  • One of the following is selected: the default reference language number corresponding to each group of Arabic numerals separated in the trademark, and the default reference language number corresponding to each individual Arabic number in the trademark.
  • the predetermined reference language number is any language number;
  • each set of other language numbers separated in the trademark the corresponding preset reference language number, corresponding to each individual language number in the trademark
  • Presetting a reference language number Presetting a reference language number, the preset reference language number being any language number
  • the constituent elements are the meaning of the symbol.
  • the minimum unit of pronunciation features includes:
  • the constituent elements are the smallest unit of the pronunciation features of Chinese characters, which are the pinyin of each Chinese character;
  • the constituent element is the smallest unit of the pronunciation feature of the graphic, and is the pinyin of the name of each thing corresponding to the coding of the trademark graphic element;
  • the constituent element is the smallest unit of the pronunciation feature of the letter, and one of the following is selected: the pronunciation of each letter combination, the pronunciation of each letter;
  • the smallest unit of the pronunciation feature whose constituent elements are numbers or symbols, one of which is selected: the pronunciation of each group of numbers separated in the trademark, the pronunciation of each individual number, the pronunciation of each group of symbols separated in the trademark, each The pronunciation of a single symbol.
  • the trademark card standard consisting of preset shape features, pronunciation features and meaning feature minimum units and various combinations thereof includes:
  • the trademark card standard consisting of multiple combinations of shape features and minimum elements of Chinese characters, including: card standard a 1 , a 2 , a 3 , a 4 , a 5 , a 6 , a 7 At least one of a 8 , a 9 , a 10 , a 11 , a 12 , a 13 , wherein
  • a 1 indicates that the combination of all the language characters and graphic element codes of the entire trademark is divided into one sub-card.
  • a 2 indicates that the combination of all the language characters and graphic element codes of the trademark as a whole in reverse order is divided into one card.
  • a 3 means that the Chinese characters in the trademarks arranged in order are divided into one card
  • a 4 indicates that the Chinese characters in the trademarks arranged in reverse order are divided into one sub-card
  • a 5 means that the Chinese characters in the trademarks are divided into one sub-cards.
  • a 6 indicates that the Chinese alphabet in the reverse order is divided into a sub-card.
  • a 7 means that each relatively independent part of the trademark is divided into one sub-card
  • a 8 means that the trademark text contains the existing Chinese word mark completely, and the part is divided into one card.
  • a 9 means that the traditional and foreign characters contained in the trademark are converted into simplified characters and then divided into one card.
  • a 10 means that each word in the trademark is replaced by a sub-card after being replaced by a near-word.
  • a 11 indicates that each adjacent Chinese character in the trademark is divided into one sub-card
  • a 12 indicates that the first Chinese character combination in the trademark is divided into one sub-card
  • a 13 means that each Chinese character in the trademark is divided into one sub-card.
  • a 1 indicates that the combination of all the language characters and graphic element codes of the trademarks arranged in order is divided into one card. That is, all the characters and graphic elements of the trademark are encoded, whether it is a combination of Chinese characters or other languages, combinations of letters, numbers, symbols or other elements, and whether or not they can form a vocabulary with common meanings.
  • the combination of all the language characters and graphic element codes of the trademark as a whole is treated as a split card.
  • Fig. 2a as an example, according to the trademark segmentation rules, it is divided into: "Gree GREE+26.1.10" sub-card
  • Figure 2c as an example, according to the standard of the trademark card, it is divided into: " ⁇ MEIXIUSHIMEI" Branch card.
  • a 2 indicates that the combination of all the language characters and graphic element codes in the reverse order of the trademark is divided into one card. That is, all the characters contained in the trademark, whether it is a combination of Chinese characters or other languages, combinations of letters, numbers, symbols or other elements, and whether or not they constitute a vocabulary with common meaning, all the trademarks are The combination of language text and graphic element coding is treated as a split card in reverse order.
  • Figure 2a as an example, according to the standard of the trademark card, it is divided into: “26.1.10+EERG Lige” card, with Figure 2c as an example, according to the standard of the trademark card: “IEMIHSUIXIEM beautiful poetry and beautiful "Sub-card.
  • the smallest unit of text is a single text, multiple words can be ordered; the smallest unit of letters, numbers, symbols is a single letter, number, symbol, multiple letters, numbers, symbol combinations can be ordered; graphic element code "26.1.
  • the 10" overall is the smallest unit of the graphic shape feature, and the numbers cannot be reordered, but the multiple graphic element codes can be changed in order (the same below).
  • a 3 indicates that the Chinese characters in the trademarks arranged in order are divided into one sub-card. That is, the Chinese characters contained in the trademark are treated as a single card in the overall order.
  • Figure 2c according to the standard of the trademark card, it is divided into: “Mixiumeimei” card.
  • a 4 indicates that the Chinese characters in the trademarks arranged in reverse order are divided into one sub-card. That is, the Chinese characters contained in the trademark are treated as a sub-card in the overall reverse order.
  • Fig. 2c as an example, according to the standard of the trademark card, it is divided into: "Beautiful poems and beautiful" card.
  • a 5 means that the Chinese characters in the trademarks arranged in order are divided into one card. That is, the Chinese number contained in the trademark is regarded as a sub-card by arranging its Chinese number and the corresponding Aber number as a whole. Taking Figure 2b as an example, it is divided into: " ⁇ " and "123" points according to the standard of the trademark card.
  • a 6 indicates that the Chinese characters in the reverse order are divided into one sub-card. That is to say, the Chinese characters contained in the trademark are regarded as a sub-card by respectively arranging the Chinese numerals and the corresponding Aber numbers in reverse order. Taking Figure 2b as an example, it is divided into: “ ⁇ ” and “321” points according to the standard of the trademark card.
  • a 7 indicates that each relatively independent part of the trademark is divided into one sub-card. That is, the relatively independent part of the trademark contains its relatively independent part as a separate card. Taking Fig. 2c as an example, according to the standard of the trademark card, it is divided into: “Mixiu”, “Shimei”, “MEIXIU SHIMEI” card. Wherein: the distinguishing rules of the relatively independent part include: different languages are divided into different relatively independent parts, and the same language type separated by symbols or spaces is combined into different relatively independent parts, and different color combinations of the same language are different. Relatively independent part.
  • a 8 means that the trademark text contains the existing Chinese word mark completely, and the part is divided into one card. That is, the trademark contains the prior Chinese word mark of the other person, and the part of the prior person's trademark is regarded as a branch card. Take Figure 2d as an example. Assume that the prior trademarks of others are: “Sitong” and “Opp”. According to the standard of the trademark card, they are divided into four parts: “Sitong” and “Opp”.
  • a 9 means that the trademark contains traditional and foreign characters converted into simplified characters and then divided into one card. That is, the trademark contains traditional and variant characters, and the traditional and foreign characters are converted into simplified characters and regarded as a split card. Taking Fig. 2e and Fig. 2f as examples, the words “ ⁇ ” and “ ⁇ ” in the trademark are respectively classified into: “ ⁇ ” participle according to the trademark card standard.
  • a 10 means that each word in the trademark is replaced by a near-word and then divided into one card. That is, the trademark contains a near-word, and the combined text of the near-word is regarded as a participle.
  • Figure 2h according to the standard of the trademark card, it is divided into: “G knives”, “Glade”, “Golding Power”, “Luo Li”, “Luo Li”, “Shovel”, “ ⁇ ” ",””power” and other participles.
  • a 11 indicates that each adjacent Chinese character in the trademark is divided into one sub-card. That is, when the number of trademark Chinese characters is three or more, each two adjacent Chinese characters in the trademark are regarded as one card. Taking Figure 2d as an example, according to the standard of the trademark card, it is divided into: “Sitong”, “Tongou” and “Opp”.
  • a 12 indicates that the first Chinese character combination in the trademark is divided into one sub-card. That is, when the number of trademark Chinese characters is three or more, the first and last Chinese characters in the trademark are regarded as one sub-card. Taking Figure 2d as an example, according to the standard of the trademark card, it is divided into: "four Pu" participle.
  • a 13 means that each Chinese character in the trademark is divided into one sub-card. Think of each Chinese character in the trademark as a split card. Taking Figure 2d as an example, according to the standard of the trademark card, it is divided into: “four”, “ ⁇ ”, “European” and “Pu”.
  • a trademark card standard consisting of a plurality of combinations of shape features and minimum elements of letters, numbers, and symbols, including sub-card standards b 1 , b 2 , b 3 , b 4 , b 5 , b 6 , At least one of b 7 , b 8 , b 9 , b 10 , b 11 , b 12 , b 13 , b 14 , wherein
  • b 1 indicates that the combination of all the language characters and graphic element codes of the trademarks arranged in order is divided into one sub-card.
  • b 2 indicates that the combination of all the language characters and graphic element codes of the trademark as a whole in reverse order is divided into one card.
  • b 4 indicates that the letter combination in the reverse order of the trademark is divided into a sub-card
  • b 5 indicates that the non-Chinese numbers contained in the listed trademarks or each individual non-Chinese number are respectively divided into one sub-cards.
  • b 6 indicates that the non-Chinese numbers contained in the reverse-ordered trademarks or each individual non-Chinese number are respectively divided into one sub-cards.
  • b 10 means that each letter in the trademark is replaced by a close letter and then divided into a minute card.
  • b 12 indicates that the letters in the trademark are arranged in different order and then divided into one sub-cards.
  • b 13 indicates that the first and last letter combination in the trademark is divided into one sub-card.
  • B 1 indicates that the combination of all the language characters and graphic element codes of the entire trademark is divided into one sub-card. That is, all the characters and graphic elements of the trademark are encoded, whether it is a combination of Chinese characters or other languages, combinations of letters, numbers, symbols or other elements, and whether or not they can form a vocabulary with common meanings.
  • the combination of all the language characters and graphic element codes of the trademark as a whole is treated as a split card. Taking Fig. 2a as an example, according to the standard of the trademark card, it is divided into: “Gree GREE+26.1.10" card, with Figure 2c as an example, according to the standard of the trademark card: "Mei Xiu Shimei MEIXIUSHIMEI "Sub-card.
  • b 3 indicates that the letter combination in the trademarks arranged in order is divided into one card. That is, the letter combination text contained in the trademark regards its overall alphabetical arrangement as one card. Taking Figure 2c as an example, it is divided into: "MEIXIUSHIMEI" card according to the trademark card standard.
  • b 5 indicates that the non-Chinese numbers contained in the sequential trademarks or each individual non-Chinese number are respectively divided into one sub-card. That is, the non-Chinese number contained in the trademark is regarded as a sub-card by arranging the non-Chinese numbers and the corresponding Aber numbers as a whole. Taking Figure 2i as an example, it is divided into: "one two three" and "123" card according to the trademark card standard.
  • b 6 indicates that the non-Chinese numbers or each individual non-Chinese number contained in the trademarks in reverse order are respectively divided into one card. That is, the non-Chinese number contained in the trademark is regarded as a sub-card by arranging its non-Chinese numbers and the corresponding Aber numbers in reverse order. Taking Fig. 2i as an example, according to the standard of the trademark card, it is divided into: "three two one" and "321" card.
  • b 7 indicates that the combination of symbols contained in the trademarks arranged in order is divided into one card. That is, the symbol combination text contained in the trademark is treated as a sub-card by arranging the symbol combination text as a whole. Taking Figure 2p as an example, it is divided into: "@" card according to the trademark card standard.
  • b 8 indicates that the combination of symbols contained in the trademarks arranged in reverse order is divided into one card. That is, the symbol combination text contained in the trademark is regarded as a sub-card by arranging the symbol combination text as a whole in reverse order. Taking Figure 2p as an example, it is divided into: "@" card according to the trademark card standard.
  • each relatively independent part of the trademark is divided into one sub-card. That is, the relatively independent part of the trademark contains its relatively independent part as a separate card. Taking Fig. 2c as an example, according to the standard of the trademark card, it is divided into: “Mixiu”, “Shimei”, “MEIXIU SHIMEI” card.
  • the distinguishing rules of the relatively independent part include: different languages are divided into different relatively independent parts, and the same language type separated by symbols or spaces is combined into different relatively independent parts, and different color combinations of the same language are different. Relatively independent part.
  • b 10 means that each letter in the trademark is replaced by a close letter and then divided into a minute card. That is, the trademark contains a near-letter letter, and the near-letter combination is regarded as a branch card. Taking Figure 2l as an example, according to the trademark card standard, it is divided into: “DC”, “DG”, “DO”, “OC”, “OO”, “OG” and other sub-cards.
  • b 11 indicates that each adjacent letter combination in the trademark is divided into one card. That is, when the number of trademark letter words is four or more, each n adjacent letters or numbers or symbols of the whole letter, number, and symbol of the trademark are regarded as a card in the original order and in the order of the first letter. Where n ranges from more than 2 to less than 50% of the total number of letters, and when the last remainder is less than one half of the preset number of n-numbers, it is merged with the previous one into a one-card, equal to Or greater than 1 ⁇ 2, independent of 1 card. Taking Fig. 2k as an example, when the value of n is 2, it is divided into: "CA”, “CAT”, “CTA”, “CAN”, “CNA” card according to the trademark card standard.
  • b 12 indicates that the letters in the trademark are arranged in different order and then divided into one card. That is, the letter combination of the whole letter combination of the trademark in the fixed order of the whole, the word and the 26 letters is used as one card and the first letter is added as one card, but the overall letter combination of the trademark has no meaning and is pressed.
  • the card formed by letter sequencing should remove duplicate letters. Taking Figure 2k as an example, according to the trademark card standard, it is divided into: “catana”, “acnt", “cacnt” card.
  • b 13 indicates that the first letter combination in the trademark is divided into one sub-card. That is, when the trademark contains letters, numbers, symbols and combination words, the first letter or number or symbol in the trademark is regarded as a branch card. Taking Figure 2k as an example, according to the trademark card standard, it is divided into: “ca” card.
  • b 14 indicates that each letter or number or symbol in the trademark is divided into one card. That is, when a trademark contains letters, numbers, symbols, and combination words, each letter or number or symbol in the trademark is treated as a single card. Taking Figure 2k as an example, according to the trademark card standard, it is divided into: “c", "a”, "t", “n” card.
  • a trademark card standard composed of a plurality of combination schemes of a minimum feature of a shape feature of a graphic, comprising: at least one of the card standard c 1 , c 2 , c 3 , and c 4 , wherein
  • c 1 indicates that the trademark graphic element code set is divided into one sub-card as a whole.
  • c 3 denotes that the trademark image feature descriptor generated by each image feature recognition method is respectively divided into one card
  • c 4 indicates that the pre-set length of the trademark image feature descriptor generated by each image feature recognition method is respectively divided into one sub-card
  • the pre-set length of the trademark image feature descriptor refers to the length of consecutively connected pixel points on the contour of the trademark image set in advance, and the consecutively connected pixels are represented by a feature string set or a digital set, and the value range is a trademark. Image feature descriptor, or 0.1%-50% of the total length of the number set.
  • c 1 indicates that the trademark graphic element code set is divided into one sub-card as a whole.
  • the trademark graphic element code of the Vienna classification standard is generally used in the trademark industry to indicate the characteristics of the trademark graphic.
  • the whole code of all graphic elements of the trademark is treated as a single card.
  • the trademark graphic element codes obtained through the search query are: 26.1.12a, 26.2.5, 29.1.12, and are classified according to the trademark card standard: "26.1.12a, 26.2.5, 29.1. 12" card.
  • c 2 indicates that each trademark graphic element code is divided into one card. That is: the code of each graphic element of the trademark is treated as a split card.
  • the trademark graphic element codes obtained through the search query are: 26.1.12a, 26.2.5, 29.1.12, and are divided into: “26.1.12a” and “26.2.5 according to the trademark card standard. ", "29.1.12" card.
  • c 3 indicates that the trademark image feature descriptor generated by each image feature recognition method is divided into a single card.
  • the whole of the trademark image feature descriptor generated by the trademark using each image feature recognition method is regarded as a split card.
  • the image feature recognition descriptor extracted by the image feature recognition method 1 (method of extracting pixel point numbers on the image contour line based on the 10 ⁇ 10 coordinate system standard) is as shown in FIG. 3 , wherein
  • the values of the trademark image feature descriptors are as follows:
  • the image feature descriptor extracted by the image feature recognition method 2 (method of extracting the pixel point number on the image contour line based on the 20 ⁇ 20 coordinate system standard) is as shown in FIG. 4 , wherein
  • the values of the trademark image feature descriptors (from small to large) are as follows:
  • c 4 indicates that the pre-set length of the trademark image feature descriptor generated by each image feature recognition method is respectively divided into one sub-card.
  • Each pre-set long trademark image feature string of the trademark image feature descriptor (or trademark image feature information) generated by the trademark using each image feature recognition method is regarded as a minute card.
  • the trademark image feature descriptor (or trademark image feature information) preset length is a continuous partial trademark image feature descriptor of a certain length range set according to a preset rule, and is expressed as a continuous local number or character set, and takes a value.
  • the range is from 0.1% to 50% of the total length of the image feature descriptor.
  • the image feature descriptor is divided into n image feature element units according to the following specific rules, and each image feature element unit is preset for one image feature descriptor:
  • the image feature descriptor is divided into several groups according to a preset segmentation length standard, and each group is regarded as one image feature element unit;
  • the last group of the above segmentation is less than 50% of the preset segmentation length, combined with the above and is an image feature element unit, equal to or more than 50%, and the remaining characters are grouped into one group, which is regarded as An image feature element unit.
  • the image feature recognition method 1 (method of extracting the pixel point digital set on the image contour line based on the 10 ⁇ 10 coordinate system standard) is used to extract the sequence ( As shown in Figure 3, the trademark image feature descriptors extracted from the method of pixel point numbers on the outline of the image are divided into the following 11 cards according to the standard of the card:
  • the image feature recognition method 1 (method of extracting the pixel point digital set on the image contour line based on the 20 ⁇ 20 coordinate system standard) is used to extract the order (along The outline of the contour line in the clockwise direction of each adjacent point)
  • the image feature descriptor extracted by the method of the pixel point number on the image contour line is as shown in Fig. 4, and is divided into the following 11 card according to the standard of the card. :
  • a trademark card standard composed of a plurality of combination schemes of minimum components of the pronunciation features of Chinese characters, including: at least one of the card standard d 1 , d 2 , and d 3 , wherein
  • d 1 indicates that the Pinyin syllable of each Chinese character in the trademark is divided into one sub-card.
  • d 2 indicates that the pinyin corresponding to the overall Chinese character in the trademark is divided into a sub-card.
  • d 3 indicates that each Chinese character in the trademark is replaced with a near-word and the Pinyin syllable is divided into one sub-card.
  • d 1 indicates that the Pinyin syllable of each Chinese character in the trademark is divided into one sub-card.
  • the Pinyin syllable of each Chinese character of the trademark is regarded as a split card.
  • the pinyin syllables of "ge” and “force” are respectively “ge” and “li”, and are respectively classified into “ge” and “li” sub-cards according to the standard of the trademark card.
  • d 2 indicates that the pinyin corresponding to the overall Chinese character in the trademark is divided into one sub-card.
  • the Pinyin syllable of the entire Chinese character of the trademark is regarded as a split card. Taking Fig. 2h as an example, the pinyin syllables of "ge” and “force” are respectively “ge” and “li”, and are respectively classified into “geli” card according to the standard of the trademark card.
  • d 3 indicates that each Chinese character in the trademark is replaced with a near-word and the Pinyin syllable is divided into one sub-card.
  • the word “grid” is replaced with the word “ ⁇ ”
  • the word “force” is replaced with the word “knife”
  • the pinyin syllables of "sickle” are "ge” and "dao” respectively.
  • the trademark card standard it is divided into: “ge dao” card.
  • a trademark card standard composed of a plurality of combination schemes of a minimum unit of pronunciation features of letters, numbers, and symbols, including: at least one of the card standard e 1 , e 2 , e 3 , and e 4 , among them,
  • e 1 means that the pronunciation syllable of each English word in the trademark is divided into a sub-card.
  • e 2 means that the whole letter combination obtained by replacing the letter combination in the trademark by the close letter combination is divided into one card
  • e 3 indicates that the pronunciation syllable of each digit in the trademark is divided into a sub-card.
  • e 4 indicates that the pronunciation syllable of each symbol in the trademark is divided into one sub-card.
  • e 1 indicates that the pronunciation syllable of each English word in the trademark is divided into one sub-card.
  • the pronunciation syllable of each English word of the trademark is regarded as a split card.
  • the pronunciation syllables of the words “one”, “two”, and “three” are respectively "[w ⁇ n]", “[tu]”, “[ ⁇ ri]”, and are respectively classified according to the trademark card standard. It is: “[w ⁇ n]”, “[tu:]”, “[ ⁇ ri:]” card.
  • e 2 indicates that the entire letter combination obtained by replacing the letter combination in the trademark by the close letter combination is divided into one card. That is, the trademark contains a close letter combination, and the near letter combination is regarded as a split card. Taking Fig. 2k as an example, in which the "CA” and “KA” pronunciations are the same or similar, they are divided into “CATANA” and “KATANA” sub-cards according to the trademark card division standard.
  • e 3 indicates that the pronunciation syllable of each digit in the trademark is divided into one sub-card.
  • the pronunciation syllable of each digit of the trademark is considered to be a split card.
  • “one”, “two”, and “three” are the pronunciation syllables of the English digital words are "[w ⁇ n]", “[tu]”, “[ ⁇ ri]”, according to the trademark card standard. Divided into: “[w ⁇ n]”, “[tu:]”, “[ ⁇ ri:]” card.
  • e 4 indicates that the pronunciation syllable of each symbol in the trademark is divided into one sub-card. That is, the trademark contains a symbol, and the pronunciation of the symbol is regarded as a minute card. Take Figure 2d as an example, where "@” is a symbol and its pronunciation is “at” or According to the standard of the trademark card, it is divided into: “at” or Branch card.
  • a trademark card standard composed of a plurality of combination schemes of minimum components of a pronunciation feature composed of a constituent element including: a card division standard f 1 , wherein f 1 represents the name of each thing corresponding to the coding of the graphic element of the trademark
  • the pinyin is divided into a split card.
  • the trademark graphic element code obtained through the search query is: 5.7.13, and the corresponding reflection of the graphic element code indicates that the name of each thing is “apple” or “persimmon”, and the pinyin is: “ Pingguo” or “shizi” is divided into: “pingguo” or “shizi” card according to the standard of the trademark card.
  • a trademark card standard composed of a plurality of combinations of meaning elements having a minimum feature unit of Chinese characters, including: at least one of the card standard g 1 , g 2 , g 3 , and g 4 , wherein
  • g 1 means that the trademark contains the existing Chinese word mark in the trademark server, and the whole trademark has no meaning.
  • the part containing the existing Chinese word mark is divided into one card.
  • g 2 indicates that the vocabulary included in the Chinese dictionary and the Chinese character combination of the existing Chinese character trademark in the trademark dictionary all match, and the matching part is respectively divided into one sub-card.
  • g 3 indicates that the Chinese vocabulary contained in the trademark is replaced by a synonym and is divided into a sub-card.
  • g 4 indicates that the trademark as a whole has no meaning, and the overall Chinese text is divided into one sub-card.
  • g 1 means that the trademark contains the existing Chinese word mark in the trademark server and the whole trademark has no meaning (the whole text cannot match the vocabulary included in the Chinese dictionary), and the part containing the existing Chinese word mark is divided into one card. .
  • the existing Chinese word mark has formed its unique meaning, which can be regarded as a unique noun, and the noun is regarded as a split card. Taking Figure 2d as an example, the “four-way opto” has no meaning in its whole. Assume that the existing Chinese-language trademarks include “Opp”, which is divided into: “Opp” sub-card according to the standard of this trademark.
  • g 2 indicates that the vocabulary included in the Chinese dictionary and the Chinese character combination of the existing Chinese character trademark in the trademark dictionary all match, and the matching portion is respectively divided into one sub-card. Taking Figure 2g as an example, it is divided into: “computer” card according to the trademark card standard.
  • g 3 indicates that the Chinese vocabulary contained in the trademark is replaced by a synonym and then divided into a sub-card. That is, the trademark contains Chinese vocabulary, and the synonym of the Chinese vocabulary is regarded as a sub-card.
  • Fig. 2g as an example, "computer” and “computer” are synonymous, and are divided into “computer” equalization cards according to the trademark card standard.
  • g 4 indicates that the trademark as a whole has no meaning, and the overall Chinese text is divided into one sub-card. That is, the overall Chinese language of the trademark has no meaning, and the overall Chinese text of the trademark is regarded as a sub-card. Taking Figure 2d as an example, the “Sitong Opus” overall Chinese has no meaning. According to the standard of the trademark card, it is divided into: “Four Opto”.
  • a trademark card standard consisting of a plurality of combinations of meaning elements of a combination of letters, numbers and symbols, including: card standard h 1 , h 2 , h 3 , h 4 , h 5 , h 6 , at least one of h 7 , h 8 , h 9 , wherein
  • h 1 indicates that the overall letter combination of the trademark is composed of a combination of words recorded in an English dictionary or other language dictionary, and the overall word combination is divided into one sub-card.
  • h 2 indicates that the trademark contains words in the English dictionary or other language dictionary, and each word is divided into one card.
  • h 3 indicates that the trademark contains words in the English dictionary or other language dictionary, and the synonyms of each word are divided into one card.
  • h 4 indicates that the overall letter combination of the trademark does not match the words recorded in the English dictionary or other language dictionary, and the overall letter combination is divided into one sub-card.
  • h 5 means that each group of numbers separated in the trademark is divided into one card.
  • h 6 indicates that the overall digital combination of the trademark is divided into one sub-card
  • h 7 indicates that the overall symbol combination of the trademark is divided into one sub-card
  • h 8 means that each symbol of the trademark is divided into one sub-card
  • h 9 means that the trademark completely contains the existing letter combination trademark of the trademark server, and the whole trademark has no meaning. The part containing the existing letter combination trademark is divided into one sub-card.
  • h 1 indicates that the overall letter combination of the trademark is composed of a combination of words recorded in an English dictionary or other language dictionary, and the overall word combination is divided into one sub-card.
  • the overall letter combination of the trademark is composed of English words, and all the words are combined and divided into one sub-card, and are divided into: “one two three” sub-card according to the standard of the trademark card.
  • h 2 indicates that the trademark contains words in the English dictionary or other language dictionary, and each word is divided into one card. That is, the trademark contains English words, and each English word is treated as a separate card. Taking Figure 2i as an example, according to the trademark card standard, it is divided into: "one", "two", "three” card.
  • h 3 indicates that the trademark contains words in the English dictionary or other language dictionary, and the synonyms of each word are divided into one card. That is, the trademark contains an English synonym, and the English synonym is regarded as a sub-card.
  • “ability” and “capacity”, “capability”, “genius”, “talent”, “competence”, “faculty”, “gift”, “aptitude”, etc. all have the ability to represent people.
  • the meaning of "capability” is divided into: “ability”, “capacity”, “capability”, “genius”, “talent”, “competence”, “faculty”, “gift”, “aptitude” according to the standard of the trademark card. A halve card.
  • h 4 indicates that the overall letter combination of the trademark does not match the words recorded in the English dictionary or other language dictionary, and the overall letter combination is divided into one sub-card, that is, the overall letter combination of the trademark is not recorded in the English dictionary or other language dictionary. word. Taking Fig. 2a as an example, "GREE" is not a word recorded in an English dictionary or other language dictionary, and is classified into "GREE” card according to the standard of the trademark card.
  • h 5 means that each group of numbers separated in the trademark is divided into one card, that is, when the number in the trademark is divided into two groups or more, each group of numbers is divided into one card. Where the numbers are separated means that the numbers in the trademark are separated by words, symbols, letters, pictures, spaces, and the like.
  • h 6 indicates that the overall digital combination of the trademark is divided into one sub-card, that is, the whole number of the trademark is combined and divided into one sub-card.
  • h 7 indicates that the overall symbol combination of the trademark is divided into one sub-card, that is, the symbols contained in the trademark are combined and divided into one sub-card.
  • h 8 means that each symbol of the trademark is divided into one sub-card, that is, each symbol contained in the trademark is divided into one sub-card.
  • h 9 means that the trademark completely contains the existing letter combination trademark of the trademark server, and the whole trademark has no meaning.
  • the part containing the existing letter combination trademark is divided into one sub-card. That is to say: the trademark completely contains the existing letter combination trademark of the trademark server and the trademark has no meaning as a whole, and the part containing the existing letter combination trademark is divided into one sub-card.
  • Figure 2a suppose the trademark completely contains the trademark letter “GREE” of the trademark server, and “GREE” is not a word recorded in an English dictionary or other language dictionary. The whole trademark has no meaning, and is cut according to the standard of the trademark card. Divided into: "GREE" card.
  • a trademark card standard consisting of a plurality of combination schemes, wherein the constituent elements are graphical meanings, and at least one of the sub-card standards i 1 and i 2 , wherein
  • i 1 indicates that the name of each thing corresponding to the trademark graphic element code is divided into one card
  • i 2 indicates that the trademark image feature descriptor corresponds to the trademark graphic element code, and each transaction name corresponding to the trademark graphic element code is divided into one sub-card.
  • i 1 indicates that the name of each thing corresponding to the trademark graphic element code is divided into one card.
  • Processing method Firstly, by establishing a transaction name dictionary file, recording the correspondence between the trademark graphic element code and the name of the object described by the trademark graphic element code, and searching for the name of the thing matching the dictionary file by using the graphic element coding of the input trademark as the retrieval condition.
  • the name of the thing is treated as the name of the thing described by the trademark image feature descriptor, and the name of the thing is treated as a split card. Taking FIG.
  • the trademark graphic element code obtained by the search query is: 5.7.13
  • the description of the trademark graphic element code is “Apple” or “Persimmon”
  • the figure describes the name of the object “Apple” and Or “persimmon” is regarded as a sub-card.
  • the name of each object corresponding to the graphic element code "5.7.13” is divided into: “Apple” and "Persimmon” card according to the standard of this card.
  • i 2 indicates that the trademark image feature descriptor corresponds to the trademark graphic element code, and each transaction name corresponding to the trademark graphic element code is divided into one sub-card.
  • the trademark graphic element corresponding to the trademark image feature descriptor is encoded and obtained by the following method for the name of each object corresponding to the trademark graphic element encoding:
  • the result mark is regarded as a graphic element code of the input mark by using the mark code of the mark pattern of the prior art; Then, by establishing a transaction dictionary file, recording the correspondence between the trademark graphic element code and the name of the object described by the trademark graphic element code; finally, finding the name of the thing matching the thing dictionary file by using the graphic element of the input trademark as the search condition,
  • the name of the thing is treated as the name of the thing described by the trademark image feature descriptor, and the name of the thing is treated as a split card.
  • the trademark graphic element code obtained by searching for the trademark image feature descriptor is: “5.7.13”, and the corresponding “thing name” is “apple” and “persimmon”.
  • the trademark image feature descriptors are respectively classified into “Apple” and "Persimmon” card according to the standard of the card.
  • Y a trademark card standard composed of a plurality of combination schemes in which the constituent elements are the minimum unit of the exception adjustment text, and includes at least one of the sub-card standards y 1 and y 2 , wherein
  • y 1 means that the trademark contains the exception adjustment text, and the exception adjustment text is divided into one sub-card as a whole;
  • y 2 means that the trademark contains the exception adjustment text, and the exception adjustment text is divided into one sub-card.
  • the exception adjustment text includes one or more of the following presets: a place name of an administrative area above the county level, a foreign place name known to the public, a generic product name, indicating the quality of the product, main raw materials, functions, uses, weights, quantities, and other characteristics. Words, generic names for goods, and texts that are significantly weaker.
  • the “Electrical Appliance” in the trademark text “Geli Electrical Appliances” is a generic commodity name. It is divided into “electrical” sub-cards according to the y 1 sub-card standard. It is divided into “electricity” according to the y 2 sub-card standard. , “device” card.
  • the content of the trademark constituent elements, the Chinese characters include: the Chinese characters and their combinations contained in the trademark.
  • the graphics include: the trademark image and the pixel information of the image.
  • the letters include: the letters and their combinations in the trademark, and the numbers or symbols include: Contains Chinese numbers, Arabic numerals and other language numbers, or symbols.
  • Figures 2a to 2p show the original trademarks of the original trademarks.
  • These trademark images may include the contents of the trademark components: Chinese characters, letters, numbers, symbols, graphics, etc.
  • the contents of the constituent elements of the input trademark are generally The identification is obtained by entering the search portal of the trademark search, and can also be obtained by image recognition or OCR character recognition.
  • the content of the constituent elements of the sample trademark is generally encoded from various trademark name data records and trademark graphic elements in the existing trademark database. The acquisition is identified in the data record.
  • the content that identifies the constituent elements of the trademark is: Chinese characters: Gree, letter: GREE, graphic: image of the trademark, trademark graphic element code: 26.1.10 (Note: from the existing trademark database The tagged information identifies the acquisition).
  • the purpose of the trademark card is to provide data support for trademark approximation evaluation, the data consists of minimum unit data of various features and combinations thereof, and the minimum unit data and its combination scheme constitute a trademark card.
  • Standard, the smallest unit of various features includes:
  • the smallest unit of shape features includes:
  • the smallest element of the shape feature of the Chinese character may be one of the following: each Chinese character, or each stroke of each Chinese character.
  • the smallest unit of the shape feature of the trademark is Chinese characters: each Chinese character contained in the trademark, namely: “Ge” and “Strength”;
  • the smallest element of the shape feature of the graphic element may be selected as one of the following: the trademark graphic element coding, and the set of pixel points on the outline of the long trademark image is preset.
  • the trademark graphic element code namely: "26.1.10";
  • the smallest feature of the shape feature that constitutes a letter is one of the following: a word for each letter combination, or each letter. Taking Fig. 2a as an example, the smallest unit of the trademark shape feature is: "GREE” when “words for each letter combination” is selected, or "G", "R”, " when "each letter” is selected. E", "E”;
  • the constituent elements are the smallest unit of shape features of Chinese numerals, and one of them is selected as follows: a combination of Chinese numbers, each individual Chinese number. Taking Figure 2b as an example, the smallest unit of the shape feature of the Chinese character is: “ ⁇ ” when “Chinese combination of numbers” is selected, and “ ⁇ ” and “ ⁇ for each individual Chinese number”. ",” " ⁇ ";
  • the smallest element of the shape feature of the Arabic numerals is selected as follows: a combination of Arabic numerals, each individual Arabic number;
  • the smallest element of the shape feature of the other language numbers is selected as follows: a combination of other language numbers, each individual other language number;
  • the constituent elements are the smallest feature of the shape feature of the symbol: for each individual symbol.
  • the constituent elements are the smallest unit of meaning of Chinese characters: when the overall Chinese character combination of a trademark is composed of a combination of words recorded in a Chinese dictionary, each word is the smallest unit of meaning features. Otherwise, the overall Chinese character combination of the trademark is the smallest unit of meaning features. ;
  • the constituent element is the meaning of the graphic.
  • the constituent elements are the meaning of the letters.
  • the overall letter combination is the smallest unit of meaning features;
  • the constituent elements are the smallest unit of meaning meaning of Chinese numerals. One of them is selected as follows: the default reference language number corresponding to each group of Chinese characters separated in the trademark, and the default reference language number corresponding to each single Chinese digit in the trademark. Wherein the predetermined reference language number is any language number;
  • the constituent element is the smallest unit of meaning of Arabic numerals.
  • One of the following is selected: the default reference language number corresponding to each group of Arabic numerals separated in the trademark, and the default reference language number corresponding to each individual Arabic number in the trademark.
  • the predetermined reference language number is any language number;
  • each set of other language numbers separated in the trademark the corresponding preset reference language number, corresponding to each individual language number in the trademark
  • Presetting a reference language number Presetting a reference language number, the preset reference language number being any language number
  • the constituent elements are the meaning of the symbol.
  • the minimum unit of pronunciation features includes:
  • the constituent elements are the smallest unit of the pronunciation features of Chinese characters, which are the pinyin of each Chinese character;
  • the constituent element is the smallest unit of the pronunciation feature of the graphic, and is the pinyin of the name of each thing corresponding to the coding of the trademark graphic element;
  • the constituent element is the smallest unit of the pronunciation feature of the letter, and one of the following is selected: the pronunciation of each letter combination, the pronunciation of each letter;
  • the smallest unit of the pronunciation feature whose constituent elements are numbers or symbols, one of which is selected: the pronunciation of each group of numbers separated in the trademark, the pronunciation of each individual number, the pronunciation of each group of symbols separated in the trademark, each The pronunciation of a single symbol.
  • the content of the sample mark is composed of Chinese characters, figures, letters, numbers or symbols, and the smallest unit of the shape feature of each component of the sample mark, and the smallest unit of the pronunciation feature are obtained.
  • the smallest unit of the meaning feature, the segmentation information of various characters and graphics generated or converted by the combination scheme of each minimum unit, and the segmentation information is used as the sample trademark card information, and each preset is established.
  • the trademark card standard pre-approximation evaluation score.
  • the preset approximation evaluation score is as shown in Table 1, wherein t 1 , t 2 , t 3 , t 4 , ... t 56 respectively represent the preset approximation evaluation scores corresponding to the respective card standards.
  • the pre-determined approximation evaluation score of the predetermined trademark sub-card standard is determined by the personnel with certain trademark review professional experience on the ranking of each trademark sub-card standard on the approximation degree of the trademark. Set the approximation evaluation score, which ranges from 0.1% to 100%.
  • step S120 the trademark card and the content are processed according to the preset trademark card standard, and the specific processing process includes: (1) establishing a minimum unit of the preset shape feature, the pronunciation feature and the meaning feature. (2) Identifying whether the input trademark is composed of Chinese characters, graphics, letters, numbers or symbols, and obtaining the constituent elements; (3) Extracting the input trademarks The shape feature minimum unit, the pronunciation feature minimum unit and the meaning feature minimum unit of the constituent elements; (4) extracting the division of various characters and graphics generated or converted by each combination scheme according to the established trademark card classification standard Information, using these segmentation information as input trademark card information.
  • the input trademark is used as the processing object, and each combination scheme is extracted for the input trademark.
  • the input trademark card information includes: a product category range and a query content
  • the “query content” is a trademark card information obtained by processing an input trademark by a trademark card, including a card type, a card content, The number of cards, the card standard used, the default score of the card standard, and so on.
  • the input trademark card information includes: U 0 , ⁇ 1 , V 0 , ⁇ 2 , M 0 , Y 0 , where U 0 indicates that the input trademark is based on the trademark card standard a 13 , b 14 , the number of cards obtained by c 2 , c 4 or a combination thereof; ⁇ 1 represents the number of cards scored or the number of words scored by the card-based standard a 13 , b 14 , c 2 , c 4 of the exception adjustment text contained in the input trademark; 0 indicates the number of cards obtained by the input trademark based on the trademark card standard d 1 , d 2 , d 3 , e 1 , e 2 , e 3 , e 4 or a combination thereof; ⁇ 2 indicates the exception adjustment text contained in the input trademark Based on the card number d 1 , d 2 , d 3 , e 1 , e 2 , e 3 , e 4 scored by the number of
  • step S130 the sample trademark card information stored in the trademark storage is searched by using the input trademark card information set as a search keyword, and the card information and the card matching information of the relevant result trademark are obtained.
  • the input trademark card information set is used as a search keyword, and includes a segmentation information of various characters and graphics according to the foregoing, as a trademark card that reflects a shape feature, a pronunciation feature, and a meaning feature of the trademark. information.
  • the distribution card information and the card matching information of the result trademark include: the registration number of the result trademark, the product category, the type of the card, the content of the card, the number of cards, the card standard adopted, and the preset score of the card standard Value, etc.
  • the card information and the card matching information of the result trademark include Y a , U a , U b , U c , V a , V b , V c , M 1 , M 2 , M 3 , M 4 , J i , n, k i , r, T i , where Y a represents the number of cards of the resulting trademark based on the trademark card standard y 1 or y 2 ; U a represents the result of the trademark removal exception adjustment text and input The trademark is based on the number of matching cards in the score card of the trademark card standard a 13 , b 14 , c 2 , c 4 or a combination thereof; U b indicates the result of the trademark removal exception adjustment
  • M 3 indicates the result of the comparison after the trademark removal exception adjustment text and the input trademark Based on the number of matching cards of the trademark card standard g 3 , M 4 indicates the result of the comparison.
  • step S140 calculating the formula of the trademark shape near rate, the trademark proximity rate, the trademark sound near rate, and the search keyword matching score rate, respectively calculating the trademark proximity rate and the trademark between the input trademark and the result trademark. Probability rate, trademark tone near rate and search keyword matching score rate.
  • W unit represents the trademark form closeness
  • U 0 represents the number of cards obtained by the input trademark based on the trademark card standard a 13 , b 14 , c 2 , c 4 or a combination thereof
  • U a represents the result of the trademark removal exception adjustment text The number of cards that match the score card of the input mark based on the trademark card standard a 13 , b 14 , c 2 , c 4 or a combination thereof
  • U b indicates the result mark removal exception adjustment text and the input mark based on the trademark card The number of matching cards in the score card of the standard a 10 , b 10 or a combination thereof
  • U c indicates that the resulting trademark and the input trademark are based on the trademark card standard a 13 , b 14 , c 2 , c 4 or a combination thereof and a 10 , b 10 or a combination of the resulting scorecards inserted into the number of mismatched cards
  • ⁇ 1 represents the exception adjustment text contained in the input trademark based on the card
  • the input trademark is “Gree” as shown in Fig. 2h.
  • the card collection of various feature types of the input trademark includes “Heli”, “G knives”, “Sickles”, “Ge”, “Strength”, “ ⁇ ” and “ ⁇ ”, and use this as a search key to search the trademark database, and obtain the relevant query result trademarks are “Geli”, “G knives” and “Sickles”, and assume that the value of ⁇ 1 is 90. %, ⁇ 2 has a value of 150%.
  • the input trademark and the result trademark do not contain the trademark exception adjustment text, and ⁇ 1 is 0.
  • the calculation result of each trademark and input trademark is calculated according to the calculation formula of the trademark shape near rate. rate:
  • V 0 represents the number of cards obtained by the input trademark based on the trademark card standard d 1 , d 2 , d 3 , e 1 , e 2 , e 3 , e 4 or a combination thereof
  • V a indicates the number of cards that match the score card of the input mark based on the trademark card standard d 1 , d 2 , e 1 , e 3 , e 4 or a combination thereof after the result mark removal exception adjustment text
  • V b indicates the result mark The number of cards that match the score card of the input mark based on the trademark card standard d 3 , e 2 or a combination thereof after removing the exception adjustment text
  • V c indicates that the result mark and the input mark are based on the mark card standard d 1 , d 2 , e 1 , e 3 , e 4 or a combination thereof and the matching scorecards obtained by d 3 , e 2 or a combination thereof are inserted into the number of
  • the number of cards scored or the number of syllables scored by the card standard d 1 , d 2 , d 3 , e 1 , e 2 , e 3 , e 4 , ⁇ 1 and ⁇ 2 are preset adjustment weights, and the value ranges are 10 Between % and 300%.
  • the input trademark is “Geli” as shown in Fig. 2h
  • the trademark memory is searched by using the card collection of various feature types of the input trademark as the search keyword
  • the relevant query result trademark is “Geli” and “G knives”.
  • " ⁇ " the syllables of the corresponding characters are "ge”, “li”, “dao”, and assume that the value of ⁇ 1 is 90%, the value of ⁇ 2 is 150%, and the trademark and result trademark are entered. No trademark exception adjustment text is included in the middle, ⁇ 2 is 0, and the sound proximity of each result trademark and input trademark is calculated according to the calculation formula of the trademark sound near rate:
  • S meaning represents the trademark proximity rate
  • M 0 represents the number of cards that match the score card of the result mark based on the trademark card classification standards g 1 , g 2 , g 3 , g 4 after inputting the trademark removal exception adjustment text
  • M 1 indicates the result of the comparison.
  • the number of matching cards after the trademark removal exception adjustment text and the input trademark based on the trademark card standard g 1 ; M 2 indicates the result of the comparison.
  • the trademark removal exception adjustment text is followed by the input trademark based on the trademark card standard g 2
  • the number of matching cards; M 3 indicates the result of the comparison.
  • the trademark removal exception is adjusted after the text is entered with the number of matching cards based on the trademark card standard g 3 , and M 4 indicates the result of the comparison.
  • the trademark removal exception is adjusted after the text is based on the input trademark.
  • the number of matching cards of the trademark card standard g 4 , ⁇ 1 , ⁇ 2 , ⁇ 3 respectively represent the adjustment parameters for M 2 , M 3 , M 4 , and the value rule is: when M 1 , M 2 , M 3 appear When two or more parameters in M 4 are not 0 at the same time, the first parameter in M 1 , M 2 , M 3 , M 4 is a valid parameter, and the rest are invalid parameters.
  • ⁇ 1 , ⁇ 2 , ⁇ 3 have a value of 0; when M 1 is 0 and M 2 is not 0, ⁇ 1 is 1, ⁇ 2 , ⁇ 3 have a value of 0; when M 1 , M 2 are 0 and M 3 is not 0, ⁇ 2 is 1, ⁇ 3 is 0; when M 1 , M 2 When M 3 is 0 and M 4 is not 0, ⁇ 3 is 1; ⁇ represents an adjustment parameter whose input trademark is different from the number of comparative trademark characters, and the value ranges from 1% to 90%.
  • the input trademark is the "four-pass op" shown in Figure 2c. It is assumed that the trademark memory is searched by using the card set of various feature types of the input trademark as the search key, and the trademark of the prior trademark is stored in the trademark memory. And the "four links" data, obtain the relevant query result trademarks are "Opp” and "four links", assuming that the value of ⁇ is 10%, according to the calculation formula of the trademark proximity rate, each result trademark and input trademark are calculated. Probability rate:
  • the input trademark is “Gree Electric” as shown in Fig. 2o. It is assumed that the trademark storage is searched by using the card collection of various feature types of the input trademark as the search key, and the trademark "meal” is stored in the trademark storage.
  • the data, the relevant query result trademark is "Gree", assuming that the value of ⁇ is 10%, the process of calculating the righteousness ratio of the resulting trademark and the input trademark according to the calculation formula of the trademark proximity rate is as follows:
  • the “electric appliance” in the input trademark is “common name of commodity service”, which belongs to “exception adjustment text” and should be removed when calculating;
  • the search keyword matching score rate includes at least one of the following: the search keyword matching comprehensive average score rate, the search keyword matching category average score rate, the search keyword matching category highest score rate, and the search keyword matching category weighted highest score.
  • S keywork represents the search keyword matching score rate
  • S 1 represents the search keyword matching comprehensive average score rate
  • S 2 represents the search keyword matching category average score rate
  • S 3 represents the search keyword matching category highest score rate
  • S 4 Indicates that the search keyword matches the classification weighted highest score rate.
  • S 1 (J 1 + J 2 + J 3 + ... + J n ) ⁇ n
  • S 1 represents a comprehensive average scoring rate of the search keyword matching
  • J 1 , J 2 , J 3 , ... J n respectively represent a preset approximation of the trademark card standard corresponding to each participle of the result trademark matching the input trademark. The score is evaluated, and n is the number of cards in which the resulting trademark matches the input trademark.
  • S 2 represents an average score of the search keyword matching classification
  • k 1 represents an average score of the predetermined proximity evaluation scores of the trademark card standard corresponding to the segmentation of the result mark and the input mark in the first feature type.
  • the value, k 2 represents the average score of the initial approximation evaluation score of the trademark card standard corresponding to each segment of the result mark and the input mark in the second feature type
  • k 3 indicates that the result mark and the input mark are in the first (3)
  • k r represents the trademark corresponding to each participle of the result trademark and the input trademark matching the r-feature type
  • the preset approximation of the card standard evaluates the average score of the score, and r represents the number of matched feature types.
  • S 3 represents the highest scoring rate of the search keyword matching classification
  • T 1 represents the highest score among the evaluation scores of the preset approximation of the word segmentation standard corresponding to the segmentation index of the input trademark and the input token in the first feature type.
  • the value T 2 represents the highest score in the evaluation score of the default approximation of the word segmentation standard corresponding to each segmentation word of the input trademark in the second feature type
  • T 3 represents the result trademark and the input trademark in the first 3
  • T r represents the segmentation word corresponding to the segmentation word of the result trademark and the input trademark matching the r feature type
  • the highest score in the evaluation score of the standard preset approximation, and r represents the number of matching feature types.
  • S 4 represents the highest score of the search keyword matching classification weighting
  • T 1 represents the highest score of the predetermined proximity evaluation score of the word segmentation standard corresponding to the segmentation of the result trademark and the input trademark in the first feature type.
  • the value T 2 represents the highest score in the initial approximation score of the word segmentation standard corresponding to the segmentation code of the input trademark in the second feature type
  • T 3 indicates that the result trademark and the input trademark are in the third.
  • Tr i indicates the word segmentation standard corresponding to each segment of the result mark and the input trademark matching the r feature type.
  • the preset approximation evaluates the highest score in the score, r denotes the number of matched feature types, and ⁇ 1 , ⁇ 2 , ⁇ 3 , ... ⁇ r respectively represent the resulting trademark and the input trademark in the first feature type, the second
  • the calculation weights of the highest scores in the predetermined approximation scores of the word segmentation criteria corresponding to each participle matching the feature type, the third feature type, ..., the r-th feature type, ⁇ 1 , ⁇ 2 , ⁇ 3 , ... ⁇ r ranges from 1% to 80%, and the total of all calculated weights is 100%.
  • the feature type is divided into a shape and a sound, including: a shape feature type (T 1 ), a pronunciation feature type (T 2 ), and a meaning feature type (T 3 ); and the component content division includes: Chinese character feature Type (T 1 ), alphabetic feature type (T 2 ), digital character feature type (T 3 ), symbolic text feature type (T 4 ), graphic element coding graphic feature type (T 5 ), image feature descriptor graphic feature Type (T 6 ).
  • the input trademark is the “Stone Tongpu” shown in Figure 2d
  • the trademark storage is searched by using the card collection of various feature types of the trademark as the retrieval keyword
  • the relevant query result trademark is “Oup” and “four links”
  • the sub-cards matched by the search keywords include the card scores obtained according to the trademark classification card of a 11 , a 12 , a 13 , e 1 , g 1 , and assume a 11 , a 12 , a 13 , e 1 , g 1 , j 1
  • the default approximation evaluation scores of each trademark card standard are 50%, 60%, 40%, 40%, 100%, shape feature type (T 1 ), pronunciation sign
  • the calculation formula calculates the result as follows:
  • the feature type includes three feature types: a shape feature type, a pronunciation feature type and a meaning feature type.
  • the card obtained according to the trademark classification card of a 11 , a 12 , a 13 belongs to the shape feature type
  • the word segment obtained according to the e 1 trademark card standard belongs to the pronunciation feature type, according to the g 1 trademark card standard.
  • the obtained participle belongs to the meaning feature type, and the matching feature type number r is 3.
  • the highest scored trademark card standard in the search keyword shape feature type is the a 12 trademark card standard, the score is 60%, and the highest score in the search keyword pronunciation feature type is the e 1 trademark.
  • the scorecard standard the score is 40%, the highest score in the search keyword meaning feature type is the g 1 trademark card standard, the score is 100%, and the matching feature type number r is 3.
  • step S150 calculating the comprehensive quantized value of the trademark approximate degree according to the calculation formula of the predetermined trademark approximate degree comprehensive quantitative value, and then sorting the result trademark by using the size of the trademark approximate degree comprehensive quantized value.
  • TM near W unit ⁇ Q 1 +S sound ⁇ Q 2 +S meaning ⁇ Q 3 +S keywork ⁇ Q 4
  • TM near represents the comprehensive quantitative value of trademark approximation
  • W unit represents the trademark proximity
  • S sound represents the trademark sound near rate
  • S meaning represents the trademark proximity rate
  • S keywork represents the search keyword matching score rate
  • Q 1 , Q 2 , Q 3 and Q 4 respectively represent the weight of the trademark near rate, the trademark sound near rate, the trademark near rate and the search keyword matching score rate
  • the Q 1 , Q 2 , Q 3 and Q 4 values range from 5 Between % and 95%, but the total of all calculated weights is 100%.
  • the input trademark is “Gree Electric” as shown in Figure 2o
  • the obtained result trademarks are “Geli” and “Yi Li”.
  • the “Electrical Appliance” of the input trademark is “Common Name of Commodity Service”, which belongs to the trademark exception adjustment text.
  • the word segmentation matched by the calculated search keyword includes a card split according to the card classification criteria of a 8 , a 11 , a 12 , a 13 , d 2 , e 1 , g 1 , and sets a 8 , a 11
  • the preset approximation evaluation scores corresponding to a 12 , a 13 , d 2 , e 1 , and g 1 are 90%, 50%, 60%, 40%, 60%, 40%, 100%, and ⁇ 1 respectively .
  • the value is 90%, the value of ⁇ 2 is 80%, the value of ⁇ 1 is 90%, and the value of ⁇ 2 is 80%.
  • the preset trademark near rate, trademark sound near rate, trademark meaning The weights of the near rate and the search keyword matching score rate are 40%, 15%, 30%, and 15%, respectively.
  • the trademark participle is divided according to the shape and sound meaning, and the feature type includes the shape feature type and the pronunciation.
  • Feature Types and Meaning Feature Types Three feature types. The highest score rate of the search keyword matching classification is taken as the search keyword matching score rate, and the “electric appliance” is the “common name of the commodity service”, which belongs to the trademark exception adjustment parameter.
  • the calculation process and results of the comprehensive quantitative value of the trademark proximity degree are as follows:
  • “Electrical appliance” is an exception adjustment text. After entering the trademark “Geli Electric Appliance”, the exception adjustment text is removed, and “Geli” is entered. After entering the trademark “Geli Electric Appliance”, the “Geli” after the exception adjustment text is removed, and the comparative result “Geli” is matched. It is a sub-card that matches the result mark based on the trademark card standard g 1 after the input mark removal exception adjustment text, and M 0 and M 1 are both 1. In the present embodiment, both M 2 and M 3 are 0, and "Geli” is not described in the Chinese dictionary, and is a meaningless combination, so M 4 is 1. The input trademark and the result trademark are not the same as the number of the adjustment parameters, and the ⁇ is 10%, then:
  • Search keyword matching score rate The calculation process of the highest score rate of the search keyword matching classification in this embodiment is as follows:
  • the highest score T 1 in the search keyword shape feature type is the trademark card standard of the trademark card standard a 8 with a score of 90%.
  • the highest score T 2 of the search keyword pronunciation feature type is the trademark score of the trademark card standard e 1 .
  • Card standard the score is 40%
  • the highest score T 3 in the search keyword meaning feature type is the trademark card standard of the trademark card standard g 1 , the score is 100%
  • the matching feature type number r is 3.
  • TM near W unit ⁇ Q 1 +S sound ⁇ Q 2 +S meaning ⁇ Q 3 +S keywork ⁇ Q 4
  • “Electrical appliance” is an exception adjustment text. After entering the trademark “Geli Electrical Appliance”, the exception adjustment text is removed, and “Geli” is entered. After entering the trademark “Geli Electrical Appliance”, the “Geli” after the exception adjustment text is removed and the comparison result trademark “ ⁇ ” matches. Is the sub-card that matches the trademark of the result trademark based on the sub-card standard g 2 after entering the trademark removal exception adjustment text, M 0 and M 2 are both 1, M 1 and M 3 are all 0, "Gree” It is not recorded in the Chinese dictionary and is a meaningless combination, so M 4 is 1. The input trademark and the result trademark are not the same as the number of the adjustment parameters, and the ⁇ is 10%, then:
  • the highest score T 1 in the search keyword shape feature type is the trademark card standard of the trademark card standard a 8 with a score of 90%.
  • the highest score T 2 of the search keyword pronunciation feature type is the trademark score of the trademark card standard e 1 .
  • Card standard the score is 40%
  • the highest score T 3 in the search keyword meaning feature type is the trademark card standard of the trademark card standard g 1 , the score is 100%
  • the matching feature type number r is 3.
  • TM near W unit ⁇ Q 1 +S sound ⁇ Q 2 +S meaning ⁇ Q 3 +S keywork ⁇ Q 4
  • Figure 5 shows a screenshot of the top 24 results trademark report interface sorted by the trademark proximity comprehensive quantized value.
  • the graphic shown in FIG. 2n is used as an input trademark
  • the product range is the 42nd category of the Nice classification
  • the registered country is China
  • the result trademark is calculated by the comprehensive quantitative value of the trademark similarity of the foregoing method of the present invention. Screenshot of the trademark report interface.
  • the approximation evaluation and sorting method of the trademark query result according to the present invention can effectively overcome the defects and drawbacks of the one-sided or missed check result of the sorting result caused by the single feature sorting method of the traditional trademark query result, and can comprehensively reflect the shape, sound and meaning of the trademark.
  • the combined comprehensive features enhance the accuracy and recall of the same or similar judgment of the trademark.
  • the comprehensive quantitative value of the trademark approximation is used to effectively quantify the visual result of the trademark image abstraction, and the quantitative evaluation level of the trademark approximation is greatly improved.
  • the invention improves the standardization level of the same or approximate judgment of the trademark, and narrows the difference between the sorting result of the trademark query result approximation ranking result and the same or similar sorting result in the sense of the trademark law expected by the examiner, and realizes the input trademark and the sample trademark. Whether it constitutes a good evaluation of the same or similar trademarks, and accelerates the progress of trademark review.
  • the invention only needs to input the trademark to be retrieved into the system once to obtain the best comprehensive sorting result, and overcomes the need for the existing trademark retrieval system to continuously perform human-computer interaction to obtain different sorting and display results, or use manual screening.
  • the resulting search results are subjectively too strong.
  • the device further relates to a device for approximating and ranking the trademark query result approximation
  • FIG. 6 is a schematic structural diagram of the trademark query result approximation evaluation and sorting device in the embodiment of the present invention, and a trademark query result.
  • the approximation evaluation and sorting device includes:
  • Sample trademark card pre-processing module used to process the trademark card image and content according to the preset trademark card standard.
  • the specific processing process includes: (1) establishing a preset shape feature, pronunciation features and Meaning-characteristics The minimum number of units of the trademark division card standard, (2), whether the sample trademark is identified by Chinese characters, graphics, letters, numbers or symbols, to obtain the content of the constituent elements, (3), The minimum feature unit of the shape characteristic of each component of the sample trademark, the smallest unit of the pronunciation feature and the smallest unit of the meaning feature; (4) extracting various characters and graphics generated or converted by each combination scheme according to the established trademark card classification standard The segmentation information, the segmentation information is used as the sample trademark card information, and the approximation evaluation score of each preset trademark card standard is set;
  • the input trademark card processing module is used for processing the trademark image and the content according to the preset trademark card standard.
  • the specific processing process includes: (1) establishing a preset shape feature, pronunciation feature and meaning (2) The identification of whether the input trademark is composed of Chinese characters, graphics, letters, numbers or symbols, and the content of the constituent elements; (3), extraction Entering the minimum feature of the shape feature of each component of the trademark, the minimum unit of the pronunciation feature and the minimum unit of the meaning feature; (4) extracting various characters and graphics generated or converted by each combination scheme according to the established trademark card classification standard The segmentation information, using the segmentation information as the input trademark card information;
  • the trademark search module searches for the sample trademark card information stored in the trademark storage by using the input trademark card information set as a search key, and obtains the card information and the card matching information of the relevant result trademark;
  • the trademark shape near rate calculation module is configured to calculate a trademark form close ratio between the input trademark and the result trademark according to a preset formula for calculating the trademark shape near rate;
  • the trademark near-rate calculation module is used to calculate the trademark proximity ratio between the input trademark and the result trademark according to the preset formula of the trademark near-probability ratio;
  • Trademark sound near rate calculation module used to calculate the trademark sound near rate between the input trademark and the result trademark according to the preset trademark sound near rate calculation formula
  • the search keyword matching score rate calculation module is configured to calculate a search keyword matching score rate between the input trademark and the result trademark according to a preset search keyword matching score rate calculation formula
  • the calculation module of the comprehensive approximation value of the trademark approximation used to calculate the comprehensive quantized value of the trademark approximation according to the preset formula of the comprehensive approximation of the trademark approximation, and then sort the result trademark by the size of the trademark approximation .
  • Embodiment 2 is a diagrammatic representation of Embodiment 1:
  • the embodiment provides a method for approximating and ranking the trademark query result, which is different from the first embodiment in that the order of the first two steps in the approximation evaluation and sorting method of the trademark query result is different, and the embodiment is specific. Includes the following steps:
  • Step S210 Performing a trademark card processing on the input trademark image and content according to a preset trademark card standard.
  • the specific processing process includes: (1) establishing a plurality of combinations of minimum shape elements, preset sound features, and meaning features.
  • the trademark sub-card standard constituted by the scheme, (2), whether the input trademark is identified by Chinese characters, graphics, letters, numbers or symbols, and the content of the constituent elements is obtained; (3) extracting the constituent elements of the input trademark Shape feature minimum unit, pronunciation feature minimum unit and meaning feature minimum unit; (4), according to the established trademark card standard, extract the segmentation information of various characters and graphics generated or converted by each combination scheme, These segmentation information is used as input trademark card information;
  • Step S220 Performing a trademark card processing on the sample trademark image and content according to a preset trademark card division standard, and the specific processing process includes: (1) establishing a plurality of combinations of minimum shape elements of preset shape features, pronunciation features, and meaning features.
  • the trademark sub-card standard constituted by the scheme, (2) whether the sample trademark is identified by Chinese characters, graphics, letters, numbers or symbols, and the content of the constituent elements is obtained, (3) the shape of each component of the sample trademark The minimum feature unit, the minimum unit of the pronunciation feature and the minimum unit of the meaning feature; (4) extracting the segmentation information of various characters and graphics generated or converted by each combination scheme according to the established trademark card classification standard, and The segmentation information is used as the sample trademark card information, and the approximation evaluation score of each preset trademark card standard is set;
  • Step S230 Searching the sample trademark card information stored in the trademark storage by using the input trademark card information set as a search keyword, and acquiring the card information and the card matching information of the relevant result trademark;
  • Step S240 Calculating a formula according to a preset trademark proximity rate, a trademark proximity rate, a trademark sound near rate, and a search keyword matching score rate, respectively calculating a trademark proximity rate and a trademark proximity rate between the input trademark and the result trademark. , trademark tone near rate and search keyword matching score rate;
  • Step S250 Calculating the approximate degree of the acquired trademark according to the calculation formula of the integrated trademark approximate degree comprehensive quantized value The quantized values are then used to rank the resulting trademarks using the approximate size of the trademark proximity.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • General Health & Medical Sciences (AREA)
  • Library & Information Science (AREA)
  • Health & Medical Sciences (AREA)
  • Multimedia (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Medical Informatics (AREA)
  • Computing Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Software Systems (AREA)
  • Image Analysis (AREA)
  • Character Discrimination (AREA)
  • Document Processing Apparatus (AREA)
  • Image Processing (AREA)
  • Editing Of Facsimile Originals (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Machine Translation (AREA)

Abstract

本发明公开一种商标查询结果近似度评价和排序方法、装置。从形、音、义的不同角度对样本商标和输入商标进行分卡处理,分别得到样本商标与输入商标的形、音、义分卡信息;通过检索获取结果商标与输入商标的匹配信息,然后按照预设的公式分别计算结果商标和输入商标之间的形近率、商标义近率、商标音近率和检索关键词匹配得分率,从而计算获取商标近似度综合量化值,并利用商标近似度综合量化值的大小对结果商标进行排序。能全面反映商标的形、音、义的综合特征近似程度,提升商标相同或近似判断的准确性和查全率。

Description

一种商标查询结果近似度评价和排序方法、装置 技术领域
本发明涉及商标信息检索领域,具体涉及一种商标查询结果近似度评价和排序方法、装置。
背景技术
商标查询对于商标注册、管理及保护具有重要的意义,其作用体现在:能及时发现商标注册申请的障碍,弄清商标能否安全使用,发现他人抢注的商标,了解商标的法律状态、查清相关商标权利范围的详细信息等。而目前的商标查询系统所报告的结果商标存在以下缺陷和弊端:
1、传统的商标查询系统所报告的结果商标的特征值虽有很多种,如:商标中文名称特征、商标英文名称特征、拼音字母特征、图形要素编码特征、图像特征描述符等,但任一种特征值都不能全面反映商标的形、音、义相结合的综合特征,从而造成商标相同或近似可能判断错误。
2、传统的商标查询系统的结果商标排序方法一般是按照其中单一特征进行排序,但两种及以上特征不能并列排序,因此,其报告和显示的排序结果商标有一定的片面性。
3、传统的商标查询系统的结果商标排序,需要与检索用户不断地交互,排序结果并不具有固定性或商标相同或近似判断的标准一致性,因而,传统方法所述的商标近似度排序与《商标法》意义上的商标相同或近似存在较大的差异。
如申请号为201410043915.0的中国发明专利,名称为:商标查询系统和方法,其中商标查询系统包括:查询模块,用于接收待查询商标;特征提取模块,用于提取所述待查询商标的商标特征;索引库,用于存储所提取到的待查询商标的商标特征;商标库,用于存储现有商标;特征库,用于存储所述现有商标的商标特征;检索模块,用于将所述待查询商标的商标特征与所述现有商标的商标特征进行匹配;以及显示模块,用于显示所述匹配的结果。通过提取待查询商标的商标特征,将所提取到的商标特征与储存在特征库中的现有商标的商标特征进行匹配,并显示匹配结果,如此降低了审查员的工作量,提高工作效率。
该专利说明书第0043段公开现有商标近似度的计算方法或实现的方法:检索模块106主要用于实现检索匹配过程,根据相关度计算方法,实现商标的匹配和筛选,最终将得到符合要求的结果反馈给用户。检索模块106基于内容的查询向用户提供检索接口,将用户的检索请求转化为可以对数据库进行操作的提问。检索允许可以针对全局对象,如整个商标,也允许针对其中的子对象以及任意组合形式来进行。检索模块106返回的结果可以按照相似程度进行排列输出,显示模块107可以显示经排序的现有商标,而且如果有必要可以基于得到的检索结果进行进一步的查询。由于基于内容的检索实现的是相似性检索,模仿人类的认知过程进行,因此,还需要在与检索用户不断地交互中提炼检索结果。
上述专利技术方案仅能解决用户的单一或逐一检索请求的匹配相似程度分别排序的问题,但无法解决由多项检索请求而产生的能匹配多项检索请求相似程度的综合排序的问题,由于现有的商标任一个单一特征不能全面反映商标的形、音、义相结合的综合特征,按单一特征的相似度排序结果不一定符合《商标法》意义上的商标相同或近似的要求,其所获的相似度排序显结果可能造成商标查询系统的用户误认为排在前的商标可能是《商标法》意义上的商标相同或近似,可能导致商标注册、管理及保护工作的严重失误。另一方面,现有的商标查询方法,对商标近似度的排序还需要与商标查询系统的用户不断地进行交互,以提供多种不同特征匹配相似程度的排序结果供用户参考,也增加了用户查询操作的工作量。
发明内容
鉴于此,本发明的目的在于提供一种商标查询结果近似度评价和排序方法、装置,能够获取对检索出的结果商标与输入商标在多特征方面进行综合评价的商标近似度综合量化值, 并按照综合量化值的大小进行对结果商标进行排序,使用户看到的结果商标更符合《商标法》意义上的商标相同或近似的要求,避免单一特征排序不能全面反映商标的多种特征从而造成商标检索的漏报错报等缺陷。
为实现上述目的,本发明的技术方案如下:
一种商标查询结果近似度评价和排序方法,对近似商标查询结果进行近似度评价和排序处理,其中包括以下步骤:
步骤S110:对样本商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对样本商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容,(3)、样本商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为样本商标分卡信息,并设定每一预设的商标分卡标准的近似度评价分值;
步骤S120:对输入商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对输入商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容;(3)、提取输入商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为输入商标分卡信息;
步骤S130:以输入商标分卡信息集合作为检索关键词对存储于商标存储器的样本商标分卡信息进行检索,获取相关的结果商标的分卡信息及分卡匹配信息;
步骤S140:按照预设的商标形近率、商标义近率、商标音近率和检索关键词匹配得分率计算公式,分别计算输入商标与结果商标之间的商标形近率、商标义近率、商标音近率和检索关键词匹配得分率;
步骤S150:按照预设的商标近似度综合量化值的计算公式,计算获取商标近似度综合量化值,然后利用商标近似度综合量化值的大小对结果商标进行排序。
所述商标查询结果近似度评价和排序方法,其中,步骤S110和步骤S120所述“形状特征最小单元、读音特征最小单元、含义特征最小单元”和“商标分卡标准”包括:
1)形状特征最小单元包括:
构成要素为汉语文字的形状特征最小单元,选如下之一:每一汉语文字,每一汉语文字的每一笔划;
构成要素为图形的形状特征最小单元,选如下之一:商标图形要素编码,预设定长的商标图像轮廓线上像素点集;
构成要素为字母的形状特征最小单元,选如下之一:每一语种的单词,每一个字母;
构成要素为汉语数字的形状特征最小单元,选如下之一:汉语数字的组合,每一单个的汉语数字;
构成要素为阿拉伯数字的形状特征最小单元,选如下之一:阿拉伯数字的组合,每一单个的阿拉伯数字;
构成要素为其他语种数字的形状特征最小单元,选如下之一:其他语种数字的组合,每一单个的其他语种数字;
构成要素为符号的形状特征最小单元:为每一单个的符号。
2)含义特征最小单元包括:
构成要素为汉语文字的含义特征最小单元:商标的整体汉语文字组合由汉语词典所记载 的词语组合构成时,每个词语为含义特征最小单元,否则,商标的整体汉语文字组合为含义特征最小单元;
构成要素为图形的含义特征最小单元:商标图形要素编码所对应的每一事物的名称;
构成要素为字母的含义特征最小单元:商标的整体字母组合由英文词典所记载的单词组合构成,或其他语种词典所记载的单词组合构成时,每个单词为含义特征最小单元,否则,商标的整体字母组合为含义特征最小单元;
构成要素为汉语数字的含义特征最小单元,选如下之一:商标中被分隔的每一组汉语数字所对应的预设基准语种数字,商标中每一单个汉语数字所对应的预设基准语种数字,其中,所述预设基准语种数字为任何语种数字;
构成要素为阿拉伯数字的含义特征最小单元,选如下之一:商标中被分隔的每一组阿拉伯数字所对应的预设基准语种数字,商标中每一单个阿拉伯数字所对应的预设基准语种数字,其中,所述预设基准语种数字为任何语种数字;
构成要素为其他语种数字的含义特征最小单元,选如下之一:商标中被分隔的每一组其他语种数字所述对应的预设基准语种数字,商标中每一单个的其它语种数字所对应的预设基准语种数字,所述预设基准语种数字为任何语种数字;
构成要素为符号的含义特征最小单元:商标中每一符号所对应的符号名称。
3)读音特征最小单元包括:
构成要素为汉语文字的读音特征最小单元,为每一汉语文字的拼音;
构成要素为图形的读音特征最小单元,为商标图形要素编码所对应的每一事物的名称的拼音;
构成要素为字母的读音特征最小单元,选如下之一:每一字母组合的读音、每一个字母的读音;
构成要素为数字或符号的读音特征最小单元,选如下之一:商标中被分隔的每一组数字的读音,每一单个数字的读音,商标中被分隔的每一组符号的读音,每一单个符号的读音。
4)商标分卡标准包括:
A、由构成要素为汉语文字的形状特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准a1、a2、a3、a4、a5、a6、a7、a8、a9、a10、a11、a12、a13中的至少一种,其中,
a1表示按顺序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡,
a2表示按逆序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡,
a3表示按顺序排列的商标中的汉语文字切分为一个分卡,
a4表示按逆序排列的商标中的汉语文字切分为一个分卡,
a5表示按顺序排列的商标中含有汉语数字切分为一个分卡,
a6表示按逆序排列的商标中含有汉语数字切分为一个分卡,
a7表示商标中每个相对独立的部分分别切分为一个分卡,
a8表示商标文字中完整包含现有汉语文字商标,将该部分切分为一个分卡,
a9表示商标中含有的繁体、异体字转换为简体字后切分为一个分卡,
a10表示商标中每个文字替换为形近字后切分为一个分卡,
a11表示商标中每相邻汉语文字分别切分为一个分卡,
a12表示商标中首尾汉字组合切分为一个分卡,
a13表示商标中每个汉语文字切分为一个分卡;
B、由构成要素为字母、数字、符号的形状特征最小单元多种组合方案所构成的商标分卡标准,,包括:分卡标准b1、b2、b3、b4、b5、b6、b7、b8、b9、b10、b11、b12、b13、b14中的至少一种,其中,
b1表示按顺序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡,
b2表示按逆序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡,
b3表示按顺序排列的商标中的字母组合切分为一个分卡,
b4表示按逆序排列的商标中的字母组合切分为一个分卡,
b5表示按顺序排列的商标中含有的非汉语数字或每一单个的非汉语数字分别切分为一个分卡,
b6表示按逆序排列的商标中含有的非汉语数字或每一单个的非汉语数字分别切分为一个分卡,
b7表示按顺序排列的商标中含有的符号组合切分为一个分卡,
b8表示按逆序排列的商标中含有的符号组合切分为一个分卡,
b9表示商标中每个相对独立部分分别切分为一个分卡,
b10表示商标中每字母被形近字母替换后切分为一个分卡,
b11表示商标中每相邻字母组合分别切分为一个分卡,
b12表示商标中字母按照不同定序排列后分别切分为一个分卡,
b13表示商标中首尾字母组合切分为一个分卡,
b14表示商标中每个字母或数字或符号分别切分为一个分卡;
C、由构成要素为图形的形状特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准c1、c2、c3、c4中的至少一种,其中,
c1表示商标图形要素编码集合整体切分为一个分卡,
c2表示每一商标图形要素编码切分为一个分卡,
c3表示每种图像特征识别方法所产生的商标图像特征描述符整体分别切分为一个分卡,
c4表示每种图像特征识别方法所产生的商标图像特征描述符预设定长分别切分为一个分卡,所述商标图像特征描述符预设定长是指预先设置的商标图像轮廓线上连续相连的像素点的长度,连续相连的像素点以特征字符串集、或数字集表示,取值范围为商标图像特征描述符、或数字集总长度的0.1%-50%;
D、由构成要素为汉语文字的读音特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准d1、d2、d3中的至少一种,其中,
d1表示商标中每个汉语文字的拼音音节切分为一个分卡,
d2表示商标中整体汉语文字对应的拼音切分为一个分卡,
d3表示商标中每个中文文字被替换为形近字后的拼音音节切分为一个分卡;
E、由构成要素为字母、数字、符号的读音特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准e1、e2、e3、e4中的至少一种,其中,
e1表示商标中每个英文单词的读音音节切分为一个分卡,
e2表示商标中字母组合被音近字母组合替换后得到的整体字母组合分别切分为一个分卡,
e3表示商标中每个数字的读音音节切分为一个分卡,
e4表示商标中每个符号的读音音节切分为一个分卡;
F、由构成要素为图形的读音特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准f1,其中,f1表示商标的图形要素编码所对应的每一事物的名称的拼音切分为一个分卡;
G、由构成要素为汉语文字的含义特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准g1、g2、g3、g4中的至少一种,其中,
g1表示商标中完整包含商标服务器中现有汉语文字商标、且商标整体无含义,将含有现有汉语文字商标的部分切分为一个分卡,
g2表示商标中与汉语词典所收录的词汇或商标服务器中现有汉语文字商标的汉字组合 全部匹配的,将匹配部分分别切分为一个分卡,
g3表示商标中含有的汉语词汇替换为近义词后分别切分为一个分卡,
g4表示商标整体无含义的,整体汉语文字切分为一个分卡;
H、由构成要素为字母、数字、符号组合的含义特征最小单元多种组合方案所构成的商标分卡标准,包括分卡标准h1、h2、h3、h4、h5、h6、h7、h8、h9中的至少一种,其中,
h1表示商标的整体字母组合由英文词典或其他语种词典所记载的单词组合构成,整体单词组合切分为一个分卡,
h2表示商标中含有英语词典或其他语种词典所收录的单词,每个单词切分为一个分卡,
h3表示商标中含有英语词典或其他语种词典所收录的单词,将每个单词的近义词切分为一个分卡,
h4表示商标的整体字母组合与英文词典或其他语种词典所记载的单词不匹配的,整体字母组合切分为一个分卡,
h5表示商标中被分隔的每一组数字切分为一个分卡,
h6表示商标的整体数字组合切分为一个分卡,
h7表示商标的整体符号组合切分为一个分卡,
h8表示商标的每一符号切分为一个分卡,
h9表示商标完整包含商标服务器现有字母组合商标、且商标整体无含义,将含有现有字母组合商标的部分切分为一个分卡;
I、构成要素为图形的含义特征最小单元多种组合方案所构成的商标分卡标准,包括分卡标准i1、i2中的至少一种,其中,
i1表示商标图形要素编码所对应的每一事物的名称切分为一个分卡,
i2表示商标图像特征描述符对应有商标图形要素编码,商标图形要素编码所对应每个事物名称切分为一个分卡。
Y、构成要素为例外调整文字的最小单元多种组合方案所构成的商标分卡标准,包括分卡标准y1、y2中的至少一种,其中,
y1表示商标中含有例外调整文字,将例外调整文字整体切分为一个分卡;
y2表示商标中含有例外调整文字,将例外调整文字每个文字分别切分为一个分卡。
优选地,所述例外调整文字包括如下预设的一种以上:县级以上行政区域地名,公众知晓的外国地名,通用商品名称,表示商品的质量、主要原料、功能、用途、重量、数量及其他特点的词语,商品服务通用名称,显著性弱的文字。所述显著性弱的文字是指自定义的一些不具备商标显著性特征的文字。本实施例中,将例外调整文字记录于基础名称词典库,包括:世界国家和地区字典表、县以上行政区域名称字典表、国外城市名称字典表、禁用词语文字词典表等。
所述商标查询结果近似度评价和排序方法,其中,步骤S120所述“输入商标分卡信息”包括:U0、β1、V0、β2、M0、Y0,其中,U0表示输入商标基于商标分卡标准a13、b14、c2、c4或其组合所得的分卡数;β1表示输入商标中含有的例外调整文字的基于分卡标准a13、b14、c2、c4所得分卡数或文字数;V0表示输入商标基于商标分卡标准d1、d2、d3、e1、e2、e3、e4或其组合所得的分卡数;β2表示输入商标中含有的例外调整文字的基于分卡标准d1、d2、d3、e1、e2、e3、e4所得分卡数或音节数;M0表示输入商标去除例外调整文字后基于商标分卡标准g1、g2、g3、g4所得的分卡数;Y0表示输入商标基于商标分卡标准y1或y2所得的分卡数;
步骤S130所述“结果商标的分卡信息及分卡匹配信息”包括Ya、Ua、Ub、Uc、Va、Vb、Vc、M1、M2、M3、M4、Ji、n、ki、r、Ti,其中,Ya表示结果商标基于商标分卡标准y1或y2所得的分卡数;Ua表示结果商标去除例外调整文字后与输入商标基于商标分卡标准a13、b14、c2、c4或其组合所得分卡中相匹配的分卡数;Ub表示结果商标去除例外调整文字后与 输入商标基于商标分卡标准a10、b10或其组合所得分卡中相匹配的分卡数;Uc表示结果商标与输入商标基于商标分卡标准a13、b14、c2、c4或其组合和a10、b10或其组合所得相匹配的分卡中间插入不匹配分卡的处数;Va表示结果商标去除例外调整文字后与输入商标基于商标分卡标准d1、d2、e1、e3、e4或其组合所得分卡中相匹配的分卡数;Vb表示结果商标去除例外调整文字后与输入商标基于商标分卡标准d3、e2或其组合所得分卡中相匹配的分卡数;Vc表示结果商标与输入商标基于商标分卡标准d1、d2、e1、e3、e4或其组合和d3、e2或其组合所得相匹配的分卡中间插入不匹配分卡的处数;M1表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g1的匹配分卡数;M2表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g2的匹配分卡数;M3表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g3的匹配分卡数,M4表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g4的匹配分卡数;Ji表示结果商标与输入商标相匹配的第i个分卡所对应的商标分卡标准的预设近似度评价分值;n表示结果商标与输入商标相匹配的分卡项目数;ki表示结果商标与输入商标在第i特征类型相匹配的各分词所对应的商标分卡标准的预设近似度评价分值的平均分值,r表示结果商标与输入商标相匹配的特征类型数;Ti表示结果商标与输入商标在第i特征类型相匹配的各分词所对应的分词标准的预设近似度的评价分值中的最高分值;
所述特征类型通过预设的分类标准对商标分卡信息进行分类所得到的分卡类别。
所述特征类型,按形音义划分包括:形状特征类型、读音特征类型、含义特征类型;按构成要素内容划分包括:汉语文字特征类型、字母文字特征类型、数字文字特征类型、符号文字特征类型、图形要素编码图形特征类型、图像特征描述符图形特征类型。
优选地,所述商标查询结果近似度评价和排序方法,其中,步骤S140所述预设的商标形近率、商标义近率、商标音近率和检索关键词匹配得分率计算公式包括:
1)商标形近率计算公式包括:
Wunit=Ua/(U01)+[Ub/(U01)]×λ1–[Uc/(U01)]×λ2
其中,Wunit表示商标形近率,λ1、λ2为预设的调整权数,λ1和λ2的取值范围均在10%至300%之间;
2)商标音近率计算公式包括:
Ssound=Va/(V02)+[Vb/(V02)]×μ1–[Vc/(V02)]×μ2
其中,Ssound表示商标音近率,μ1、μ2为预设的调整权数,取值范围均在10%至300%之间;
3)商标义近率计算公式包括:
Smeaning=[(M1+M2×α1+M3×α2+M4×α3)/M0]-θ,
其中,Smeaning表示商标义近率,α1、α2、α3分别表示对M2、M3、M4的调整参数,取值规则:当出现M1、M2、M3、M4中二个及以上参数同时不为0时,在M1、M2、M3、M4中排列最前参数为有效参数,其余为无效参数,当M1不为0时,α1、α2、α3的值为0;当M1为0且M2不为0时,α1为1,α2、α3的值为0;当M1、M2为0且M3不为0时,α2为1,α3为0;当M1、M2、M3为0且M4不为0时,α3为1;θ表示输入商标与比较的结果商标文字数不相同的调整参数,取值范围在1%至90%之间;
4)检索关键词匹配得分率计算公式包括如下至少一项:检索关键词匹配综合平均得分率,检索关键词匹配分类平均得分率,检索关键词匹配分类最高得分率,检索关键词匹配分类加权最高得分率,即:
Skeywork=S1或Skeywork=S2或Skeywork=S3或Skeywork=S4
其中,Skeywork表示检索关键词匹配得分率,S1表示检索关键词匹配综合平均得分率,S2表示检索关键词匹配分类平均得分率,S3表示检索关键词匹配分类最高得分率,S4表示 检索关键词匹配分类加权最高得分率;
其计算公式分别为:
S1=(J1+J2+J3+……+Jn)÷n
S2=(k1+k2+k3+…+kr)÷r
S3=(T1+T2+T3+…+Tr)÷r
S4=T1×ω1+T2×ω2+T3×ω3+…+Tr×ωr
其中,ω1、ω2、ω3、……ωi分别表示结果商标与输入商标在第1特征类型、第2特征类型、第3特征类型、……、第i特征类型相匹配的各分卡所对应的分卡标准的预设近似度评价分值中的最高分值的计算权数,ω1、ω2、ω3、……ωi的取值范围在1%至80%之间,全部计算权数的合计为100%。
进一步优选地,所述商标查询结果近似度评价和排序方法,其中,步骤S150所述商标近似度综合量化值的计算公式包括:
TMnear=Wunit×Q1+Ssound×Q2+Smeaning×Q3+Skeywork×Q4
其中,TMnear表示商标近似度综合量化值,Wunit表示商标形近率,Ssound表示商标音近率,Smeaning表示商标义近率,Skeywork表示检索关键词匹配得分率,Q1、Q2、Q3、Q4分别表示商标形近率、商标音近率、商标义近率和检索关键词匹配得分率的权重数,Q1、Q2、Q3、Q4取值范围在5%至95%之间,全部计算权数的合计为100%。
另一方面,本发明还提供一种商标查询结果近似度评价和排序处理的装置,包括:
样本商标分卡预处理模块:用于对样本商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对样本商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容,(3)、样本商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为样本商标分卡信息,并设定每一预设的商标分卡标准的近似度评价分值;
输入商标分卡处理模块:用于对输入商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对输入商标是否由汉语文字,图形,字母、数字或符号构成要素进行识别,获取构成要素的内容;(3)、提取输入商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为输入商标分卡信息;
商标检索模块:以输入商标分卡信息集合作为检索关键词对存储于商标存储器的样本商标分卡信息进行检索,获取相关的结果商标的分卡信息及分卡匹配信息;
商标形近率计算模块:用于按照预设的商标形近率计算公式,计算输入商标与结果商标之间的商标形近率;
商标义近率计算模块:用于按照预设的商标义近率计算公式,计算输入商标与结果商标之间的商标义近率;
商标音近率计算模块:用于按照预设的商标音近率计算公式,计算输入商标与结果商标之间的商标音近率;
检索关键词匹配得分率计算模块:用于按照预设的检索关键词匹配得分率计算公式,计算输入商标与结果商标之间的检索关键词匹配得分率;
商标近似度综合量化值的计算模块:用于按照预设的商标近似度综合量化值的计算公式,计算获取商标近似度综合量化值,然后利用商标近似度综合量化值的大小对结果商标进 行排序。
有益效果:本发明利用预设的商标分卡标准分别从不同角度对输入商标进行切分获取形状特征、读音特征和含义特征最小单元及其组合,并计算结果商标与输入商标之间的检索关键词匹配得分率、形近率、音近率和义近率,获取商标近似度综合量化值并按照近似度综合量化值的大小进行排序,能全面反映商标的形、音、义的综合特征近似程度,提升商标相同或近似判断的准确性和查全率;采用商标近似度综合量化值,有效地将商标图像抽象的视觉结果进行量化,大幅提高商标近似度的数量化评价水平;提高了商标相同或近似判断的标准化水平,缩小了商标查询结果近似度排序结果与审查人员预期的《商标法》意义上的商标相同或近似的排序结果的差异,实现输入商标与样本商标是否构成商标相同或近似的较好评价,加速商标审查工作的进步。且本发明只需将待检索的商标一次性输入系统即可得到最佳的综合排序结果,克服现有商标检索系统需要不断地进行人机交互,以获取不同的排序和显示结果,或利用人工筛选而造成的检索结果主观性太强的缺陷。
附图说明
图1是本发明实施例1的商标查询结果近似度评价和排序方法流程示意图。
图2是本发明实施例1的示例性商标原图。
图3是图2n苹果图形商标采用10×10坐标系标准获取商标图像轮廓线上像素点的图像特征描述符图。
图4是图2n苹果图形商标采用20×20坐标系标准获取商标图像轮廓线上像素点的图像特征描述符图。
图5是本发明实施例1中采用商标近似度综合量化值排序的前24件结果商标报告界面截图。
图6是本发明实施例1的商标查询结果近似度评价和排序装置结构示意图。
图7是本发明实施例2的商标查询结果近似度评价和排序方法流程示意图。
具体实施方式
为使本发明的目的、技术方案及优点更加清楚、明确,一下参照附图并和具体实施例对本发明进一步详细说明。
实施例一
如图1所示,一种商标查询结果近似度评价和排序方法,包括以下步骤:
步骤S110:对样本商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对样本商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容,(3)、样本商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为样本商标分卡信息,并设定每一预设的商标分卡标准的近似度评价分值;
步骤S120:对输入商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对输入商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容;(3)、提取输入商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为输入商标分卡信息;
步骤S130:以输入商标分卡信息集合作为检索关键词对存储于商标存储器的样本商标分卡信息进行检索,获取相关的结果商标的分卡信息及分卡匹配信息;
步骤S140:按照预设的商标形近率、商标义近率、商标音近率和检索关键词匹配得分率计算公式,分别计算输入商标与结果商标之间的商标形近率、商标义近率、商标音近率和检索关键词匹配得分率;
步骤S150:按照预设的商标近似度综合量化值的计算公式,计算获取商标近似度综合量化值,然后利用商标近似度综合量化值的大小对结果商标进行排序。
以下基于具体实施例对上述各步骤进行具体说明,需要强调的是,本实施例为了便于理解,设置了第一、第二、第三、第四、第五步骤,实际应用中,可以根据需要对各个步骤之间的顺序进行调整。
第一,步骤S110:对样本商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对样本商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容,(3)、样本商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为样本商标分卡信息,并设定每一预设的商标分卡标准的近似度评价分值。
(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准:
两商标是否构成近似,一般可从两商标是否在形状、含义、读音方面是否存在共同性进行判断,如何找出两商标的共同性及共同成份的比例是本发明实施例需要解决的技术问题。因此,本发明实施例通过对商标在形状、含义、读音方面最小的构成单元进行细分及最小单元的组合建立商标分卡标准,可以在商标查询结果近似评价和排序过程中获得有益的技术效果。
对商标在形状、含义、读音方面最小的构成单元进行细分,包括:
1)形状特征最小单元包括:
构成要素为汉语文字的形状特征最小单元,选如下之一:每一汉语文字,每一汉语文字的每一笔划;
构成要素为图形的形状特征最小单元,选如下之一:商标图形要素编码,预设定长的商标图像轮廓线上像素点集;
构成要素为字母的形状特征最小单元,选如下之一:每一语种的单词,每一个字母;
构成要素为汉语数字的形状特征最小单元,选如下之一:汉语数字的组合,每一单个的汉语数字;
构成要素为阿拉伯数字的形状特征最小单元,选如下之一:阿拉伯数字的组合,每一单个的阿拉伯数字;
构成要素为其他语种数字的形状特征最小单元,选如下之一:其他语种数字的组合,每一单个的其他语种数字;
构成要素为符号的形状特征最小单元:为每一单个的符号。
2)含义特征最小单元包括:
构成要素为汉语文字的含义特征最小单元:商标的整体汉语文字组合由汉语词典所记载的词语组合构成时,每个词语为含义特征最小单元,否则,商标的整体汉语文字组合为含义特征最小单元;
构成要素为图形的含义特征最小单元:商标图形要素编码所对应的每一事物的名称;
构成要素为字母的含义特征最小单元:商标的整体字母组合由英文词典所记载的单词组合构成,或其他语种词典所记载的单词组合构成时,每个单词为含义特征最小单元,否则,商标的整体字母组合为含义特征最小单元;
构成要素为汉语数字的含义特征最小单元,选如下之一:商标中被分隔的每一组汉语数字所对应的预设基准语种数字,商标中每一单个汉语数字所对应的预设基准语种数字,其中,所述预设基准语种数字为任何语种数字;
构成要素为阿拉伯数字的含义特征最小单元,选如下之一:商标中被分隔的每一组阿拉伯数字所对应的预设基准语种数字,商标中每一单个阿拉伯数字所对应的预设基准语种数字,其中,所述预设基准语种数字为任何语种数字;
构成要素为其他语种数字的含义特征最小单元,选如下之一:商标中被分隔的每一组其他语种数字所述对应的预设基准语种数字,商标中每一单个的其它语种数字所对应的预设基准语种数字,所述预设基准语种数字为任何语种数字;
构成要素为符号的含义特征最小单元:商标中每一符号所对应的符号名称。
3)读音特征最小单元包括:
构成要素为汉语文字的读音特征最小单元,为每一汉语文字的拼音;
构成要素为图形的读音特征最小单元,为商标图形要素编码所对应的每一事物的名称的拼音;
构成要素为字母的读音特征最小单元,选如下之一:每一字母组合的读音、每一个字母的读音;
构成要素为数字或符号的读音特征最小单元,选如下之一:商标中被分隔的每一组数字的读音,每一单个数字的读音,商标中被分隔的每一组符号的读音,每一单个符号的读音。
由预设的形状特征、读音特征和含义特征最小单元及其多种组合方案所构成的商标分卡标准包括:
A、由构成要素为汉语文字的形状特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准a1、a2、a3、a4、a5、a6、a7、a8、a9、a10、a11、a12、a13中的至少一种,其中,
a1表示按顺序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡,
a2表示按逆序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡,
a3表示按顺序排列的商标中的汉语文字切分为一个分卡,
a4表示按逆序排列的商标中的汉语文字切分为一个分卡,
a5表示按顺序排列的商标中含有汉语数字切分为一个分卡,
a6表示按逆序排列的商标中含有汉语数字切分为一个分卡,
a7表示商标中每个相对独立的部分分别切分为一个分卡,
a8表示商标文字中完整包含现有汉语文字商标,将该部分切分为一个分卡,
a9表示商标中含有的繁体、异体字转换为简体字后切分为一个分卡,
a10表示商标中每个文字替换为形近字后切分为一个分卡,
a11表示商标中每相邻汉语文字分别切分为一个分卡,
a12表示商标中首尾汉字组合切分为一个分卡,
a13表示商标中每个汉语文字切分为一个分卡。
如下结合图2中各种商标图样说明本商标分词规则的处理方法:
a1表示按顺序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡。即商标含有的全部文字及图形要素编码,不管是汉语汉字或其他语种的文字、字母组合、数字组合、符号或其他要素之间的组合,也不管其是否能构成一个有常用含义的词汇,均将商标整体所有语种文字及图形要素编码的组合按顺序排列视为一个分卡。以图2a为示例,按照本商标分词规则切分为:“格力GREE+26.1.10”分卡,以图2c为示例,按照本商标分卡标准切分为:“美秀·诗美MEIXIUSHIMEI”分卡。
a2表示按逆序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡。即商标含有的全部文字,不管是汉语汉字或其他语种的文字、字母组合、数字组合、符号或其 他要素之间的组合,也不管其是否能构成一个有常用含义的词汇,均将商标整体所有语种文字及图形要素编码的组合逆序排列视为一个分卡。以图2a为示例,按照本商标分卡标准切分为:“26.1.10+EERG力格”分卡,以图2c为示例,按照本商标分卡标准切分为:“IEMIHSUIXIEM美诗·秀美”分卡。其中:文字的最小单元为单个文字,多文字可换序;字母、数字、符号的最小单元为单个字母、数字、符号,多个字母、数字、符号组合可换序;图形要素编码“26.1.10”整体为图形形状特征最小单元,不可将其数字再换序,但多个图形要素编码之间可换序(下同)。
a3表示按顺序排列的商标中的汉语文字切分为一个分卡。即商标含有的汉语汉字将其整体顺序排列视为一个分卡。以图2c为示例,按照本商标分卡标准切分为:“美秀诗美”分卡。
a4表示按逆序排列的商标中的汉语文字切分为一个分卡。即商标含有的汉语汉字将其整体逆序排列视为一个分卡。以图2c为示例,按照本商标分卡标准切分为:“美诗秀美”分卡。
a5表示按顺序排列的商标中含有汉语数字切分为一个分卡。即商标含有的汉语数字将其汉语数字和对应的阿伯数字整体顺序排列分别视为一个分卡。以图2b为示例,按照本商标分卡标准切分为:“壹贰叁”、“123”分卡。
a6表示按逆序排列的商标中含有汉语数字切分为一个分卡。即商标含有的汉语数字将其汉语数字和对应的阿伯数字整体逆序排列分别视为一个分卡。以图2b为示例,按照本商标分卡标准切分为:“叁贰壹”、“321”分卡。
a7表示商标中每个相对独立的部分分别切分为一个分卡。即商标含有的相对独立部分将其相对独立部分分别视为一个分卡。以图2c为示例,按照本商标分卡标准切分为:“美秀”、“诗美”、“MEIXIU SHIMEI”分卡。其中:相对独立部分的区分规则包括:不同语种区分为不同的相对独立部分,用符号或空格分隔开的同一语种文字组合为不同的相对独立部分,不同的颜色的同一语种文字组合为不同的相对独立部分。
a8表示商标文字中完整包含现有汉语文字商标,将该部分切分为一个分卡。即商标含有在先的他人汉语文字商标,将该在先的他人商标的该部分视为一个分卡。以图2d为示例,假设在先的他人商标有:“四通”、“欧普”,按照本商标分卡标准切分为:“四通”、“欧普”分词。
a9表示商标中含有繁体、异体字转换为简体字后切分为一个分卡。即商标含有繁体、异体字,将该繁体、异体字转换为简体字后视为一个分卡。以图2e、图2f为示例,按照本商标分卡标准分别将商标中的“匯”、“滙”字切分为:“汇”分词。
a10表示商标中每个文字替换为形近字后切分为一个分卡。即商标含有形近字,将该形近字的组合文字视为一个分词。以图2h为示例,按照本商标分卡标准分别切分为:“格刀”、“格刃”、“烙力”、“洛力”、“络力”、“恪刀”、“恪力”、“辂力”等分词。
a11表示商标中每相邻汉语文字分别切分为一个分卡。即当商标汉字字数在三个及以上时,将商标中相邻的每两个汉字视为一个分卡。以图2d为示例,按照本商标分卡标准分别切分为:“四通”、“通欧”、“欧普”分词。
a12表示商标中首尾汉字组合切分为一个分卡。即当商标汉字字数在三个及以上时,将商标中首尾汉字视为一个分卡。以图2d为示例,按照本商标分卡标准分别切分为:“四普”分词。
a13表示商标中每个汉语文字切分为一个分卡。即将商标中每一个汉字视为一个分卡。以图2d为示例,按照本商标分卡标准分别切分为:“四”、“通”、“欧”、“普”分词。
B、由构成要素为字母、数字、符号的形状特征最小单元多种组合方案所构成的商标分卡标准,包括分卡标准b1、b2、b3、b4、b5、b6、b7、b8、b9、b10、b11、b12、b13、b14中的至少一种,其中,
b1表示按顺序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡,
b2表示按逆序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡,
b3表示按顺序排列的商标中的字母组合切分为一个分卡,
b4表示按逆序排列的商标中的字母组合切分为一个分卡,
b5表示按顺序排列的商标中含有的非汉语数字或每一单个的非汉语数字分别切分为一个分卡,
b6表示按逆序排列的商标中含有的非汉语数字或每一单个的非汉语数字分别切分为一个分卡,
b7表示按顺序排列的商标中含有的符号组合切分为一个分卡,
b8表示按逆序排列的商标中含有的符号组合切分为一个分卡,
b9表示商标中每个相对独立部分分别切分为一个分卡,
b10表示商标中每字母被形近字母替换后切分为一个分卡,
b11表示商标中每相邻字母组合分别切分为一个分卡,
b12表示商标中字母按照不同定序排列后分别切分为一个分卡,
b13表示商标中首尾字母组合切分为一个分卡,
b14表示商标中每个字母或数字或符号分别切分为一个分卡。
如下结合图2中各种商标图样说明本商标分词规则的处理方法:
b1表示按顺序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡。即商标含有的全部文字及图形要素编码,不管是汉语汉字或其他语种的文字、字母组合、数字组合、符号或其他要素之间的组合,也不管其是否能构成一个有常用含义的词汇,均将商标整体所有语种文字及图形要素编码的组合按顺序排列视为一个分卡。以图2a为示例,按照本商标分卡标准切分为:“格力GREE+26.1.10”分卡,以图2c为示例,按照本商标分卡标准切分为:“美秀·诗美MEIXIUSHIMEI”分卡。
b2表示按逆序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡。即商标含有的全部文字,不管是汉语汉字或其他语种的文字、字母组合、数字组合、符号或其他要素之间的组合,也不管其是否能构成一个有常用含义的词汇,均将商标整体所有语种文字及图形要素编码的组合逆序排列视为一个分卡。以图2a为示例,按照本商标分卡标准切分为:“26.1.10+EERG力格”分卡,以图2c为示例,按照本商标分卡标准切分为:“IEMIHSUIXIEM美诗·秀美”分卡。
b3表示按顺序排列的商标中的字母组合切分为一个分卡。即商标含有的字母组合文字将其整体字母顺序排列视为一个分卡。以图2c为示例,按照本商标分卡标准切分为:“MEIXIUSHIMEI”分卡。
b4表示按逆序排列的商标中的字母组合切分为一个分卡。即商标含有的字母组合文字将其整体字母逆序排列视为一个分卡。以图2c为示例,按照本商标分卡标准切分为:“IEMIHSUIXIEM”分卡。
b5表示按顺序排列的商标中含有的非汉语数字或每一单个的非汉语数字分别切分为一个分卡。即商标含有的非汉语数字将其非汉语数字和对应的阿伯数字整体顺序排列分别视为一个分卡。以图2i为示例,按照本商标分卡标准切分为:“one two three”、“123”分卡。
b6表示按逆序排列的商标中含有的非汉语数字或每一单个的非汉语数字分别切分为一个分卡。即商标含有的非汉语数字将其非汉语数字和对应的阿伯数字整体逆序排列分别视为一个分卡。以图2i为示例,按照本商标分卡标准切分为:“three two one”、“321”分卡。
b7表示按顺序排列的商标中含有的符号组合切分为一个分卡。即商标含有的符号组合文字将其符号组合文字整体顺序排列分别视为一个分卡。以图2p为示例,按照本商标分卡标准切分为:“@”分卡。
b8表示按逆序排列的商标中含有的符号组合切分为一个分卡。即商标含有的符号组合文 字将其符号组合文字整体逆序排列分别视为一个分卡。以图2p为示例,按照本商标分卡标准切分为:“@”分卡。
b9表示商标中每个相对独立部分分别切分为一个分卡。即商标含有的相对独立部分将其相对独立部分分别视为一个分卡。以图2c为示例,按照本商标分卡标准切分为:“美秀”、“诗美”、“MEIXIU SHIMEI”分卡。其中:相对独立部分的区分规则包括:不同语种区分为不同的相对独立部分,用符号或空格分隔开的同一语种文字组合为不同的相对独立部分,不同的颜色的同一语种文字组合为不同的相对独立部分。
b10表示商标中每字母被形近字母替换后切分为一个分卡。即商标含有形近字母,将该形近字母组合视为一个分卡。以图2l为示例,按照本商标分卡标准分别切分为:“DC”、“DG”、“DO”、“OC”、“OO”、“OG”等分卡。
b11表示商标中每相邻字母组合分别切分为一个分卡。即当商标字母字数在四个及以上时,将商标整段字母、数字、符号的每n个相邻的字母或数字或符号按原序和定序再加首字母算视为一个分卡。其中n的取值范围在大于2小于总字母数的50%的范围,当最后1个余数少于预设的n值字母数的1半时,与前一分卡合并为一分卡,等于或大于1半时,独立为1个分卡。以图2k为示例,n的取值为2时,按照本商标分卡标准分别切分为:“CA”、“CAT”、“CTA”、“CAN”、“CNA”分卡。
b12表示商标中字母按照不同定序排列后分别切分为一个分卡。即商标整体字母组合按整体、单词和26个字母的固定顺序排序规则所分别形成的字母组合作为1个分卡和再添加首字母作为1个分卡,但商标整体字母组合整体无含义且按字母定序所形成的分卡应去除重复字母。以图2k为示例,按照本商标分卡标准分别切分为:“catana”、“acnt”、“cacnt”分卡。
b13表示商标中首尾字母组合切分为一个分卡。即当商标含有字母、数字、符号及组合词汇时,将商标中首尾字母或数字或符号视为一个分卡。以图2k为示例,按照本商标分卡标准分别切分为:“ca”分卡。
b14表示商标中每个字母或数字或符号分别切分为一个分卡。即当商标含有字母、数字、符号及组合词汇时,将商标中每一个字母或数字或符号视为一个分卡。以图2k为示例,按照本商标分卡标准分别切分为:“c”、“a”、“t”、“n”分卡。
C、由构成要素为图形的形状特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准c1、c2、c3、c4中的至少一种,其中,
c1表示商标图形要素编码集合整体切分为一个分卡,
c2表示每一商标图形要素编码切分为一个分卡,
c3表示每种图像特征识别方法所产生的商标图像特征描述符整体分别切分为一个分卡,
c4表示每种图像特征识别方法所产生的商标图像特征描述符预设定长分别切分为一个分卡,
所述商标图像特征描述符预设定长是指预先设置的商标图像轮廓线上连续相连的像素点的长度,连续相连的像素点以特征字符串集、或数字集表示,取值范围为商标图像特征描述符、或数字集总长度的0.1%-50%。
如下结合图2中各种商标图样说明本商标分卡标准的处理方法:
c1表示商标图形要素编码集合整体切分为一个分卡。即:目前在商标行业内一般采用维也纳分类标准的商标图形要素编码表示商标图形的特征。将商标所有图形要素编码整体视为一个分卡。以图2m为示例,经检索查询获得的商标图形要素编码是:26.1.12a、26.2.5、29.1.12,按照本商标分卡标准切分为:“26.1.12a,26.2.5,29.1.12”分卡。
c2表示每一商标图形要素编码切分为一个分卡。即:将商标每一个图形要素编码视为一个分卡。以图2m为示例,经检索查询获得的商标图形要素编码是:26.1.12a、26.2.5、29.1.12,按照本商标分卡标准分别切分为:“26.1.12a”、“26.2.5”、“29.1.12”分卡。
c3表示每种图像特征识别方法所产生的商标图像特征描述符整体分别切分为一个分卡。即将商标采用每一种图像特征识别方法所产生的商标图像特征描述符的整体视为一个分卡。以图2n为示例,采用图像特征识别方法一(基于10×10的坐标系标准提取图像轮廓线上像素点数字集的方法)所提取的商标图像特征描述符如图3所示,其中,定序(自小至大)商标图像特征描述符的值如下:
6,7,15,16,17,25,26,27,
22,23,24,25,26,27,28,29,31,32,39,41,48,49,51,58,61,68,69,71,79,80,81,82,89,92,93,94,95,96,97,98,99。
顺序(沿轮廓线顺时针方向逐个相邻点的顺序)商标图像特征描述符的值如下:
6,7,17,27,26,25,15,16,
22,23,24,25,26,27,28,29,39,49,48,58,68,69,79,80,79,89,99,98,97,96,95,94,93,92,82,81,71,61,51,41,31,32。
按照本分卡标准分别切分为如下2个分卡:
“6,7,15,16,17,25,26,27;22,23,24,25,26,27,28,29,31,32,39,41,48,49,51,58,61,68,69,71,79,80,81,82,89,92,93,94,95,96,97,98,99”;
“6,7,17,27,26,25,15,16;22,23,24,25,26,27,28,29,39,49,48,58,68,69,79,80,79,89,99,98,97,96,95,94,93,92,82,81,71,61,51,41,31,32”。
再以图2n为示例,采用图像特征识别方法二(基于20×20的坐标系标准提取图像轮廓线上像素点数字集的方法)所提取的图像特征描述符如图4所示,其中,定序(自小至大)商标图像特征描述符的值如下:
12,13,14,31,32,34,50,51,53,54,70,73,90,91,92,93,110,111;
85,86,87,88,93,94,95,96,103,104,105,108,109,110,111,112,113,116,117,122,123,137,138,141,142,156,157,161,176,181,196,201,216,221,236,241,256,257,261,277,278,281,282,298,302,318,322,323,337,338,343,357,363,364,369,370,375,376,384,385,386,387,388,390,391,392,393,394,395。
顺序(沿轮廓线顺时针方向逐个相邻点的顺序)商标图像特征描述符的值如下:
12,13,14,34,54,53,73,93,92,91,111,110,90,70,50,51,31,32;
85,86,87,88,108,109,110,111,112,113,93,94,95,116,117,137,138,157,156,176,196,216,236,256,257,277,278,298,318,338,337,357,376,375,395,394,393,392,391,390,370,369,388,387,386,385,384,364,363,344,343,323,322,302,282,281,261,241,221,201,181,161,141,142,122,123,103,104,105。
按照本分卡标准分别切分为如下2个分卡:
“12,13,14,31,32,34,50,51,53,54,70,73,90,91,92,93,110,111;85,86,87,88,93,94,95,96,103,104,105,108,109,110,111,112,113,116,117,122,123,137,138,141,142,156,157,161,176,181,196,201,216,221,236,241,256,257,261,277,278,281,282,298,302,318,322,323,337,338,343,357,363,364,369,370,375,376,384,385,386,387,388,390,391,392,393,394,395”;
“12,13,14,34,54,53,73,93,92,91,111,110,90,70,50,51,31,32;85,86,87,88,108,109,110,111,112,113,93,94,95,116,117,137,138,157, 156,176,196,216,236,256,257,277,278,298,318,338,337,357,376,375,395,394,393,392,391,390,370,369,388,387,386,385,384,364,363,344,343,323,322,302,282,281,261,241,221,201,181,161,141,142,122,123,103,104,105”。
c4表示每种图像特征识别方法所产生的商标图像特征描述符预设定长分别切分为一个分卡。即将商标采用每一种图像特征识别方法所产生的商标图像特征描述符(或商标图像特征信息)的每一预设定长商标图像特征字符串视为一个分卡。
所述商标图像特征描述符(或商标图像特征信息)预设定长是按预设规则设置的一定长度范围的连续的局部商标图像特征描述符,表现为连续的局部数字或字符集,取值范围在图像特征描述符总长度的0.1%至50%。本实施例中,所述图像特征描述符按如下具体规则切分为n个图像特征要素单元,每个图像特征要素单元为一个图像特征描述符预设定长:
1)按获取图像特征描述符的不同坐标系标准分别预设的切分长度,预设的切分长度的取值范围在10至100字符之间;
2)当图像特征描述符的总数小于等于预设的切分长度时,不切分,整体视为一个图像特征要素单元;
3)图像特征描述符总数大于预设的切分长度时,以预设的切分长度为标准将图像特征描述符切分为若干个分组,每一分组视为一个图像特征要素单元;
4)具体连通域特征的图像特征描述符的一部分视为一个图像特征要素单元;
5)以上切分的最后一组不足预设的切分长度的50%的,与上组合并为一个图像特征要素单元,等于或超过50%的,将剩余的字符分为一组,视为一个图像特征要素单元。
再以图2n为示例,假如预设定长的值取5组数字,采用图像特征识别方法一(基于10×10的坐标系标准提取图像轮廓线上像素点数字集的方法)提取定序(自小至大)图像轮廓线上像素点数字集的方法所提取的商标图像特征描述符如图3所示,按照本分卡标准分别切分为如下11个分卡:
“6,7,15,16,17,25,26,27”、“22,23,24,25,26,27,28,29,31,32,39,41,48,49,51,58,61,68,69,71,79,80,81,82,89,92,93,94,95,96,97,98,99”;
“6,7,15,16,17”、“25,26,27”、“22,23,24,25,26”、“27,28,29,31,32”、“39,41,48,49,51”、“58,61,68,69,71”、“79,80,81,82,89”、“92,93,94,95,96”、“97,98,99”。
再以图2n为示例,假如预设定长的值取5组数字,采用图像特征识别方法一(基于20×20的坐标系标准提取图像轮廓线上像素点数字集的方法)提取顺序(沿轮廓线顺时针方向逐个相邻点的顺序)图像轮廓线上像素点数字集的方法所提取的商标图像特征描述符如图4所示,按照本分卡标准分别切分为如下11个分卡:
“6,7,17,27,26,25,15,16”、“22,23,24,25,26,27,28,29,39,49,48,58,68,69,79,80,79,89,99,98,97,96,95,94,93,92,82,81,71,61,51,41,31,32”;
“6,7,17,27,26”、“25,15,16”、“22,23,24,25,26”、“27,28,29,39,49”、“48,58,68,69,79”、“80,79,89,99,98”、“97,96,95,94,93”、“92,82,81,71,61”、“51,41,31,32”。
D、由构成要素为汉语文字的读音特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准d1、d2、d3中的至少一种,其中,
d1表示商标中每个汉语文字的拼音音节切分为一个分卡,
d2表示商标中整体汉语文字对应的拼音切分为一个分卡,
d3表示商标中每个中文文字被替换为形近字后的拼音音节切分为一个分卡。
如下结合图2中各种商标图样说明本商标分卡标准的处理方法:
d1表示商标中每个汉语文字的拼音音节切分为一个分卡。即将商标每个汉字的拼音音节视为一个分卡。以图2h为示例,“格”和“力”的拼音音节分别为“ge”和“li”,按照本商标分卡标准分别切分为:“ge”、“li”分卡。
d2表示商标中整体汉语文字对应的拼音切分为一个分卡。即将商标整体汉字的拼音音节视为一个分卡。以图2h为示例,“格”和“力”的拼音音节分别为“ge”和“li”,按照本商标分卡标准分别切分为:“geli”分卡。
d3表示商标中每个中文文字被替换为形近字后的拼音音节切分为一个分卡。以图2h为示例,“格”文字被替换为“挌”形近字,“力”文字被替换为“刀”形近字,“挌刀”的拼音音节分别为“ge”和“dao”,按照本商标分卡标准分别切分为:“ge dao”分卡。
E、由构成要素为字母、数字、符号的读音特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准e1、e2、e3、e4中的至少一种,其中,
e1表示商标中每个英文单词的读音音节切分为一个分卡,
e2表示商标中字母组合被音近字母组合替换后得到的整体字母组合分别切分为一个分卡,
e3表示商标中每个数字的读音音节切分为一个分卡,
e4表示商标中每个符号的读音音节切分为一个分卡。
如下结合图2中各种商标图样说明本商标分卡标准的处理方法:
e1表示商标中每个英文单词的读音音节切分为一个分卡。即将商标每个英文单词的读音音节视为一个分卡。以图2i为示例,“one”、“two”、“three”单词的读音音节分别为“[wΛn]”、“[tu]”、“[θri]”,按照本商标分卡标准分别切分为:“[wΛn]”、“[tu:]”、“[θri:]”分卡。
e2表示商标中字母组合被音近字母组合替换后得到的整体字母组合分别切分为一个分卡。即商标含有音近字母组合,将该音近字母组合视为一个分卡。以图2k为示例,其中“CA”与“KA”读音相同或相近,按照本商标分卡标准切分为:“CATANA”、“KATANA”等分卡。
e3表示商标中每个数字的读音音节切分为一个分卡。即将商标每个数字的读音音节视为一个分卡。以图2i为示例,“one”、“two”、“three”为英文数字单词的读音音节分别为“[wΛn]”、“[tu]”、“[θri]”,按照本商标分卡标准分别切分为:“[wΛn]”、“[tu:]”、“[θri:]”分卡。
e4表示商标中每个符号的读音音节切分为一个分卡。即商标含有符号,将该符号的读音视为一个分卡。以图2d为示例,其中“@”为符号,其读音为“at”或
Figure PCTCN2017100187-appb-000001
按照本商标分卡标准切分为:“at”或
Figure PCTCN2017100187-appb-000002
分卡。
F、由构成要素为图形的读音特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准f1,其中,f1表示商标的图形要素编码所对应的每一事物的名称的拼音切分为一个分卡。
以图2n为示例,经检索查询获得的商标图形要素编码是:5.7.13,该图形要素编码所对应的反映描述每一事物的名称是“苹果”或“柿子”,其拼音分别为:“pingguo”或“shizi”,按照本商标分卡标准切分为:“pingguo”或“shizi”分卡。
G、由构成要素为汉语文字的含义特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准g1、g2、g3、g4中的至少一种,其中,
g1表示商标中完整包含商标服务器中现有汉语文字商标、且商标整体无含义,将含有现有汉语文字商标的部分切分为一个分卡,
g2表示商标中与汉语词典所收录的词汇或商标服务器中现有汉语文字商标的汉字组合全部匹配的,将匹配部分分别切分为一个分卡,
g3表示商标中含有的汉语词汇替换为近义词后分别切分为一个分卡,
g4表示商标整体无含义的,整体汉语文字切分为一个分卡。
如下结合图2中各种商标图样说明本商标分词规则的处理方法:
g1表示商标中完整包含商标服务器中现有汉语文字商标且商标整体无含义(整体文字不能与汉语词典所收录的词汇相匹配),将含有现有汉语文字商标的部分切分为一个分卡。该现有汉语文字商标已形成其特有的含义,可视其为一个特有的名词,将该名词视为一个分卡。以图2d为示例,“四通欧普”整体无含义,假设现有汉语文字商标中有:“欧普”,按照本商标分卡标准切分为:“欧普”分卡。
g2表示商标中与汉语词典所收录的词汇或商标服务器中现有汉语文字商标的汉字组合全部匹配的,将匹配部分分别切分为一个分卡。以图2g为示例,按照本商标分卡标准分别切分为:“电脑”分卡。
g3表示商标中含有的汉语词汇替换为近义词后分别切分为一个分卡。即商标含有汉语词汇,将该汉语词汇的近义词视为一个分卡。以图2g为示例,“电脑”与“计算机”等为近义词,按照本商标分卡标准分别切分为:“计算机”等分卡。
g4表示商标整体无含义的,整体汉语文字切分为一个分卡。即商标整体汉语无含义,将商标整体汉语文字视为一个分卡。以图2d为示例,“四通欧普”整体汉语无含义,按照本商标分卡标准切分为:“四通欧普”分卡。
H、由构成要素为字母、数字、符号组合的含义特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准h1、h2、h3、h4、h5、h6、h7、h8、h9中的至少一种,其中,
h1表示商标的整体字母组合由英文词典或其他语种词典所记载的单词组合构成,整体单词组合切分为一个分卡,
h2表示商标中含有英语词典或其他语种词典所收录的单词,每个单词切分为一个分卡,
h3表示商标中含有英语词典或其他语种词典所收录的单词,将每个单词的近义词切分为一个分卡,
h4表示商标的整体字母组合与英文词典或其他语种词典所记载的单词不匹配的,整体字母组合切分为一个分卡,
h5表示商标中被分隔的每一组数字切分为一个分卡,
h6表示商标的整体数字组合切分为一个分卡,
h7表示商标的整体符号组合切分为一个分卡,
h8表示商标的每一符号切分为一个分卡,
h9表示商标完整包含商标服务器现有字母组合商标、且商标整体无含义,将含有现有字母组合商标的部分切分为一个分卡。
如下结合图2中各种商标图样说明本商标分卡标准的处理方法:
h1表示商标的整体字母组合由英文词典或其他语种词典所记载的单词组合构成,整体单词组合切分为一个分卡。以图2i为示例,该商标的整体字母组合有英文单词构成的,将所有的单词组合在一起切分为一个分卡,按照本商标分卡标准切分为:“one two three”分卡。
h2表示商标中含有英语词典或其他语种词典所收录的单词,每个单词切分为一个分卡。即商标含有英语单词,将每个英语单词分别视为一个分卡。以图2i为示例,按照本商标分卡标准分别切分为:“one”、“two”、“three”分卡。
h3表示商标中含有英语词典或其他语种词典所收录的单词,将每个单词的近义词切分为一个分卡。即商标含有英语近义词,将该英语近义词视为一个分卡。以图2j为示例,“ability”与“capacity”、“capability”、“genius”、“talent”、“competence”、“faculty”、“gift”、“aptitude”等均具有表示人的“能力,才能”之意,按照本商标分卡标准切分为:“ability”、“capacity”、“capability”、“genius”、“talent”、“competence”、“faculty”、“gift”、“aptitude”等分卡。
h4表示商标的整体字母组合与英文词典或其他语种词典所记载的单词不匹配的,整体字 母组合切分为一个分卡,即该商标的整体字母组合不是英文词典或其他语种词典所记载的单词。以图2a为例,“GREE”不是英文词典或其他语种词典所记载的单词,按照本商标分卡标准切分为:“GREE”分卡。
h5表示商标中被分隔的每一组数字切分为一个分卡,即:商标中的数字被分隔为二组或以上组数字时,每一组数字切分为一个分卡。其中,数字被分隔是指商标中的数字之间被文字、符号、字母、图片、空格等分开。
h6表示商标的整体数字组合切分为一个分卡,即:将商标中含有的数字整体组合后切分为一个分卡。
h7表示商标的整体符号组合切分为一个分卡,即:将商标中含有的符号整体组合后切分为一个分卡。
h8表示商标的每一符号切分为一个分卡,即:将商标中含有的每一符号分别切分为一个分卡。
h9表示商标完整包含商标服务器现有字母组合商标、且商标整体无含义,将含有现有字母组合商标的部分切分为一个分卡。即:商标完整包含商标服务器现有字母组合商标且商标整体无含义,将含有现有字母组合商标的部分切分为一个分卡。以图2a为例,假设该商标完整包含商标服务器现有字母组合商标“GREE”,且“GREE”不是英文词典或其他语种词典所记载的单词,商标整体无含义,按照本商标分卡标准切分为:“GREE”分卡。
I、由构成要素为图形的含义特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准i1、i2中的至少一种,其中,
i1表示商标图形要素编码所对应的每一事物的名称切分为一个分卡,
i2表示商标图像特征描述符对应有商标图形要素编码,商标图形要素编码所对应每个事物名称切分为一个分卡。
如下结合图2中各种商标图样说明本商标分卡标准的处理方法:
i1表示商标图形要素编码所对应的每一事物的名称切分为一个分卡。处理方法:首先,通过建立事物名称词典文件,记录商标图形要素编码与该商标图形要素编码所描述事物名称的对应关系,以输入商标的图形要素编码为检索条件找出事物词典文件匹配的事物名称,该事物名称视为商标图像特征描述符所描述事物名称,该事物名称视为一个分卡。以图2n为示例,经检索查询获得的商标图形要素编码是:5.7.13,该商标图形要素编码所描述事物是“苹果”和或“柿子”,该图形所描述的事物名称“苹果”和或“柿子”视为一个分卡,图形要素编码“5.7.13”所对应每个事物名称按照本分卡标准分别切分为:“苹果”、“柿子”分卡。
i2表示商标图像特征描述符对应有商标图形要素编码,商标图形要素编码所对应每个事物名称切分为一个分卡。
商标图像特征描述符所对应的商标图形要素编码并由商标图形要素编码所对应每个事物名称通过如下方法获取:
首先,以输入商标的商标图像特征描述符作为检索关键词获取检索匹配率最高的一个结果商标之后,将该结果商标运用现有技术已标记的商标图形要素编码视为输入商标的图形要素编码;然后,通过建立事物词典文件,记录商标图形要素编码与该商标图形要素编码所描述事物名称的对应关系;最后,以输入商标的图形要素编码为检索条件找出事物词典文件匹配的事物名称,该事物名称视为商标图像特征描述符所描述事物名称,该事物名称视为一个分卡。以图2n为示例,经商标图像特征描述符(或商标图像特征信息)检索查询获得的商标图形要素编码是:“5.7.13”,对应的“事物名称”是“苹果”和“柿子”,该商标图像特征描述符按照本分卡标准分别切分为:“苹果”、“柿子”分卡。
Y、由构成要素为例外调整文字的最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准y1、y2中的至少一种,其中,
y1表示商标中含有例外调整文字,将例外调整文字整体切分为一个分卡;
y2表示商标中含有例外调整文字,将例外调整文字每个文字分别切分为一个分卡。
所述例外调整文字包括如下预设的一种以上:县级以上行政区域地名,公众知晓的外国地名,通用商品名称,表示商品的质量、主要原料、功能、用途、重量、数量及其他特点的词语,商品服务通用名称,显著性弱的文字。
以图2o为示例,商标文字“格力电器”中“电器”是通用商品名称,按照y1分卡标准切分为:“电器”分卡,按照y2分卡标准切分为:“电”、“器”分卡。
(2)、对样本商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容。
商标构成要素的内容,汉语文字包括:商标中含有的汉语文字及其组合,图形包括:商标图样图片及图片的像素信息,字母包括:商标中含有的字母及其组合,数字或符号包括:商标中含有的汉语数字、阿拉伯数字和其他语种数字、或各种符号。
图2a至图2p所示为随机给出的示例性商标原图,这些商标图像可能包括商标构成要素的内容有:汉语文字、字母、数字、符号、图形等,输入商标的构成要素的内容一般通过在商标检索的检索入口录入而识别获取,也可通过图像识别或OCR文字识别获取,样本商标的构成要素的内容一般从现有的商标数据库中的各种商标名称数据记录和商标图形要素编码数据记录中识别获取。
以图2a为例,识别获取该商标构成要素的内容是:汉语文字:格力,字母:GREE,图形:商标的图像,商标图形要素编码:26.1.10(注:从现有的商标数据库中已标记的信息识别获取)。
(3)、提取样本商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元。
在本发明的实施例中,商标分卡的目是为商标近似度评价提供数据支持,这些数据由各种特征的最小单元数据及其组合构成,最小单元数据及其组合方案构成了商标分卡标准,各种特征的最小单元包括:
形状特征最小单元包括:
构成要素为汉语文字的形状特征最小单元可选如下之一:每一汉语文字,或每一汉语文字的每一笔划。以图2a为例,商标为汉语文字的形状特征最小单元是:商标中含有的每一汉语文字,即:“格”和“力”;
构成要素为图形的形状特征最小单元可选如下之一:商标图形要素编码,预设定长的商标图像轮廓线上像素点集。以图2a为例,商标为图形的形状特征最小单元是:商标图形要素编码,即:“26.1.10”;
构成要素为字母的形状特征最小单元可选如下之一:每一字母组合的单词,或每一个字母。以图2a为例,商标为字母形状特征的最小单元是:选“每一字母组合的单词”时为“GREE”,或选“每一个字母”时为:“G”、“R”、“E”、“E”;
构成要素为汉语数字的形状特征最小单元,选如下之一:汉语数字的组合,每一单个的汉语数字。以图2b为例,商标为汉语数字的形状特征最小单元是:选“汉语数字的组合”时为“壹贰叁”,选“每一单个的汉语数字”时为:“壹”、“贰”、“叁”;
构成要素为阿拉伯数字的形状特征最小单元,选如下之一:阿拉伯数字的组合,每一单个的阿拉伯数字;
构成要素为其他语种数字的形状特征最小单元,选如下之一:其他语种数字的组合,每一单个的其他语种数字;
构成要素为符号的形状特征最小单元:为每一单个的符号。
含义特征最小单元包括:
构成要素为汉语文字的含义特征最小单元:商标的整体汉语文字组合由汉语词典所记载的词语组合构成时,每个词语为含义特征最小单元,否则,商标的整体汉语文字组合为含义特征最小单元;
构成要素为图形的含义特征最小单元:商标图形要素编码所对应的每一事物的名称;
构成要素为字母的含义特征最小单元:商标的整体字母组合由英文词典所记载的单词组合构成,或其他语种词典所记载的单词组合构成时,每个单词为含义特征最小单元,否则,商标的整体字母组合为含义特征最小单元;
构成要素为汉语数字的含义特征最小单元,选如下之一:商标中被分隔的每一组汉语数字所对应的预设基准语种数字,商标中每一单个汉语数字所对应的预设基准语种数字,其中,所述预设基准语种数字为任何语种数字;
构成要素为阿拉伯数字的含义特征最小单元,选如下之一:商标中被分隔的每一组阿拉伯数字所对应的预设基准语种数字,商标中每一单个阿拉伯数字所对应的预设基准语种数字,其中,所述预设基准语种数字为任何语种数字;
构成要素为其他语种数字的含义特征最小单元,选如下之一:商标中被分隔的每一组其他语种数字所述对应的预设基准语种数字,商标中每一单个的其它语种数字所对应的预设基准语种数字,所述预设基准语种数字为任何语种数字;
构成要素为符号的含义特征最小单元:商标中每一符号所对应的符号名称。
读音特征最小单元包括:
构成要素为汉语文字的读音特征最小单元,为每一汉语文字的拼音;
构成要素为图形的读音特征最小单元,为商标图形要素编码所对应的每一事物的名称的拼音;
构成要素为字母的读音特征最小单元,选如下之一:每一字母组合的读音、每一个字母的读音;
构成要素为数字或符号的读音特征最小单元,选如下之一:商标中被分隔的每一组数字的读音,每一单个数字的读音,商标中被分隔的每一组符号的读音,每一单个符号的读音。
(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为样本商标分卡信息,并设定每一预设的商标分卡标准的近似度评价分值。
根据前述已建立的商标分卡标准,获取样本商标由汉语文字、图形、字母、数字或符号等构成要素的内容,提取样本商标的各构成要素的形状特征的最小单元、读音特征的最小单元和含义特征的最小单元,可以将每一最小单元的组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为样本商标分卡信息,并建立每一预设的商标分卡标准预设近似度评价分值。
所述的预设近似度评价分值如表1所示,其中,t1、t2、t3、t4、……t56分别表示各个分卡标准对应的预设近似度评价分值,在本实施例中,所述对预设的商标分卡标准预设近似度评价分值,是由有一定商标审查专业经验的人员针对每一商标分卡标准对商标近似程度影响排列高低而预设近似度评价分值,取值范围在0.1%至100%之间。
表1:各个分卡标准的预设近似度评价分值
Figure PCTCN2017100187-appb-000003
Figure PCTCN2017100187-appb-000004
Figure PCTCN2017100187-appb-000005
按照前述方法得到各种的商标分卡信息,将这些分卡信息作为商标在形、音、义方面的商标近似度评价的基础数据,为解决商标检索的结果商标与输入商标的近似度评价提供有效的数据支持。
第二,步骤S120:对输入商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对输入商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容;(3)、提取输入商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为输入商标分卡信息。
在本发明实施例中,参照前述“对样本商标图像及内容按预设的商标分卡标准进行商标分卡处理”的处理过程,以输入商标为处理对象,对输入商标提取每一组合方案所生成或转换得到的各种文字、图形的切分信息。
将这些切分信息作为输入商标分卡信息。
所述的输入商标分卡信息包括:商品类别范围和查询内容,所述的“查询内容”是由输入商标经商标分卡处理而获得的商标分卡信息,包括分卡类型,分卡内容,分卡数,采用的分卡标准,该分卡标准的预设分值等。作为优选的实施方式,所述输入商标分卡信息包括:U0、β1、V0、β2、M0、Y0,其中,U0表示输入商标基于商标分卡标准a13、b14、c2、c4或其组合所得的分卡数;β1表示输入商标中含有的例外调整文字的基于分卡标准a13、b14、c2、c4所得分卡数或文字数;V0表示输入商标基于商标分卡标准d1、d2、d3、e1、e2、e3、e4或其组合所得的分卡数;β2表示输入商标中含有的例外调整文字的基于分卡标准d1、d2、d3、e1、e2、e3、e4所得分卡数或音节数;M0表示输入商标去除例外调整文字后与结果商标基于商标分卡标准g1、g2、g3、g4所得分卡中相匹配的分卡数;Y0表示输入商标基于商标分卡标准y1或y2所得的分卡数。
第三,步骤S130:以输入商标分卡信息集合作为检索关键词对存储于商标存储器的样本商标分卡信息进行检索,获取相关的结果商标的分卡信息及分卡匹配信息。
在本实施例中,所述输入商标分卡信息集合作为检索关键词,包括根据前述的各种文字、图形的切分信息,作为反映商标的形状特征、读音特征和含义特征的商标分卡的信息。
所述结果商标的分卡信息及分卡匹配信息包括:结果商标的注册号及商品类别、分卡类型,分卡内容,分卡数,采用的分卡标准,该分卡标准的预设分值等。本实施例中,所述结果商标的分卡信息及分卡匹配信息包括Ya、Ua、Ub、Uc、Va、Vb、Vc、M1、M2、M3、M4、Ji、n、ki、r、Ti,其中,Ya表示结果商标基于商标分卡标准y1或y2所得的分卡数;Ua表示结果商标去除例外调整文字后与输入商标基于商标分卡标准a13、b14、c2、c4或其组合所得分卡中相匹配的分卡数;Ub表示结果商标去除例外调整文字后与输入商标基于商标分卡标准a10、b10或其组合所得分卡中相匹配的分卡数;Uc表示结果商标与输入商标基于商标分卡标准a13、b14、c2、c4或其组合和a10、b10或其组合所得相匹配的分卡中间插入不匹配分卡的处数;Va表示结果商标去除例外调整文字后与输入商标基于商标分卡标准d1、d2、e1、e3、e4或其组合所得分卡中相匹配的分卡数;Vb表示结果商标去除例外调整文字后与输入商标基于商标分卡标准d3、e2或其组合所得分卡中相匹配的分卡数;Vc表示结果商标与输入商标基于商标分卡标准d1、d2、e1、e3、e4或其组合和d3、e2或其组合所得相匹配的分卡中间插入不匹配分卡的处数;M1表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g1的匹配分卡数;M2表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g2的匹配分卡数;M3表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g3的匹配分卡数,M4表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g4的匹配分卡数;Ji表示结果商标与输入商标相匹配的第i个分卡所对应的商标分卡标准的预设近似度评价分值;n表示结果商标与输入商标相匹配的分卡项目数;ki表示结果商标与输入商标在第i特征类型相匹配的各分词所对应的商标分卡标准的预设近似度评价分值的平均分值,r表示结果商标与输入商标相匹配的特征类型数;Ti表示结果商标与输入商标在第i特征类型相匹配的各分词所对应的分词标准的预设近似度的评价分值中的最高分值。
第四,步骤S140:按照预设的商标形近率、商标义近率、商标音近率和检索关键词匹配得分率计算公式,分别计算输入商标与结果商标之间的商标形近率、商标义近率、商标音近率和检索关键词匹配得分率。
结合具体实施例对计算公式及计算方法说明如下:
(1)商标形近率计算公式为:Wunit=Ua/(U01)+[Ub/(U01)]×λ1–[Uc/(U01)]×λ2
其中,Wunit表示商标形近率,U0表示输入商标基于商标分卡标准a13、b14、c2、c4或其组合所得的分卡数;Ua表示结果商标去除例外调整文字后与输入商标基于商标分卡标准a13、b14、c2、c4或其组合所得分卡中相匹配的分卡数;Ub表示结果商标去除例外调整文字后与输入商标基于商标分卡标准a10、b10或其组合所得分卡中相匹配的分卡数;Uc表示结果商标与输入商标基于商标分卡标准a13、b14、c2、c4或其组合和a10、b10或其组合所得相匹配的分卡中间插入不匹配分卡的处数;β1表示输入商标中含有的例外调整文字的基于分卡标准a13、b14、c2、c4所得分卡数或文字数;λ1、λ2为预设的调整权数,λ1和λ2的取值范围均在10%至300%之间。
例如:输入商标是图2h所示的“格力”,输入商标的多种特征类型的分卡集合中包括“和力”、“格刀”、“恪刀”、“格”、“力”、“恪”、“刀”,并以此作为检索关键词对商标数据库进行检索,获取相关的查询结果商标是“格力”“格刀”、“恪刀”,并假设λ1的取值为90%,λ2的取值为150%,输入商标与结果商标中均不含商标例外调整文字,β1为0,按照所述商标形近率的计算公式计算各个结果商标与输入商标的形近率:
1)输入商标“格力”与结果商标“格力”的商标形近率为:
Wunit=Ua/(U01)+[Ub/(U01)]×λ1–[Uc/(U01)]×λ2=2/(2-0)+[0/(2-0)]×90%–[0/(2-0)]×150%=1=100%。
2)输入商标“格力”与结果商标“格刀”的商标形近率为:
Wunit=Ua/(U01)+[Ub/(U01)]×λ1–[Uc/(U01)]×λ2=1/(2-0)+[1/(2-0)]×90%–[0/(2-0)]×150%=95%。
3)输入商标“格力”与结果商标“恪刀”的商标形近率为:
Wunit=Ua/(U01)+[Ub/(U01)]×λ1–[Uc/(U01)]×λ2=0/(2-0)+[2/(2-0)]×90%–[0/(2-0)]×150%=90%。
(2)商标音近率计算公式为:Ssound=Va/(V02)+[Vb/(V02)]×μ1–[Vc/(V02)]×μ2
其中,Ssound表示商标音近率,V0表示输入商标基于商标分卡标准d1、d2、d3、e1、e2、e3、e4或其组合所得的分卡数,Va表示结果商标去除例外调整文字后与输入商标基于商标分卡标准d1、d2、e1、e3、e4或其组合所得分卡中相匹配的分卡数;Vb表示结果商标去除例外调整文字后与输入商标基于商标分卡标准d3、e2或其组合所得分卡中相匹配的分卡数;Vc表示结果商标与输入商标基于商标分卡标准d1、d2、e1、e3、e4或其组合和d3、e2或其组合所得相匹配的分卡中间插入不匹配分卡的处数;β2表示输入商标中含有的例外调整文字的基于分卡标准d1、d2、d3、e1、e2、e3、e4所得分卡数或音节数,μ1、μ2为预设的调整权数,取值范围均在10%至300%之间。
例如:输入商标是图2h所示的“格力”,以输入商标的多种特征类型的分卡集合为检索关键词对商标存储器进行检索,获取相关的查询结果商标是“格力”“格刀”、“挌历”,其对应文字的音节分别是“ge”、“li”、“dao”,并假设μ1的取值为90%,μ2的取值为150%,输入商标与结果商标中均不含商标例外调整文字,β2为0,按照所述商标音近率的计算公式计算各个结果商标与输入商标的音近率:
1)输入商标“格力”(“ge”、“li”)与结果商标“格力”(“ge”、“li”)的商标音近率为:
Ssound=Va/(V02)+[Vb/(V02)]×μ1–[Vc/(V02)]×μ2=2/(2-0)+[0(2-0)]×90%–[0/(2-0)]×150%=100%。
2)输入商标“格力”(“ge”、“li”)与结果商标“格刀”(“ge”、“dao”)的商标音近率为:
Ssound=Va/(V02)+[Vb/(V02)]×μ1–[Vc/(V02)]×μ2=1/(2-0)+[0/(2-0)]×90%–[0/(2-0)]×150%=50%。
3)输入商标“格力”(“ge”、“li”)与结果商标“挌历”(“ge”、“li”)的商标音近率为:
Ssound=Va/(V02)+[Vb/(V02)]×μ1–[Vc/(V02)]×μ2=0/(2-0)+[2/(2-0)]×90%–[0/(2-0)]×150%=90%。
(3)商标义近率计算公式为:Smeaning=(M1+M2×α1+M3×α2+M4×α3)/(M0)-θ,
其中,Smeaning表示商标义近率,M0表示输入商标去除例外调整文字后与结果商标基于商标分卡标准g1、g2、g3、g4所得分卡中相匹配的分卡数,M1表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g1的匹配分卡数;M2表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g2的匹配分卡数;M3表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g3的匹配分卡数,M4表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g4的匹配分卡数,α1、α2、α3分别表示对M2、M3、M4的调整参数,取值规则为:当出现M1、M2、M3、M4中二个及以上参数同时不为0时,在M1、M2、M3、M4中排列最前参数为有效参数,其余为无效参数,当M1不为0时,α1、α2、α3的值为0;当M1为0且M2不为0时,α1为1,α2、α3的值为0;当M1、M2为0且M3不为0时,α2为1,α3为0;当M1、M2、M3为0且M4不为0时,α3为1;θ表示输入商标与比较的结果商标文字数不相同的调整参数,取值范围在1%至90%之间。
例如:输入商标是图2c所示的“四通欧普”,假设以输入商标的多种特征类型的分卡集合作为检索关键词对商标存储器进行检索,商标存储器中存储有在先商标的“欧普”和“四通”数据,获取相关的查询结果商标是“欧普”和“四通”,假设θ的取值为10%,按照所述商标义近率的计算公式计算各个结果商标与输入商标的义近率:
1)输入商标“四通欧普”与结果商标“欧普”的商标义近率:
输入商标中没有“例外调整文字”,输入商标“四通欧普”与结果商标“欧普”基于商标分卡标准g1的分卡数为1,M0和M1值均为1,输入商标“四通欧普”与商标分卡标准g2、g3、g4不适用,M2、M3、M4值均为0,输入商标“四通欧普”与比较的结果商标“欧普”基于商标分卡标准g4的分卡数为1,M4值均为1,假设θ的取值为10%,计算结果如下:
Smeaning=[(M1+M2×α1+M3×α2+M4×α3)/M0]-θ=[(1+0+0+1×0)/1]-10%=90%。
2)输入商标“四通欧普”与比较的结果商标“四通”的商标义近率:
输入商标没有“例外调整文字”,输入商标“四通欧普”与比较的结果商标“四通”基于商标分卡标准g1的分卡数为1,M0和M1值均为1,输入商标“四通欧普”与商标分卡标准g2、g3、g4不适用,M2、M3、M4值均为0,输入商标“四通欧普”与比较的结果商标“四通”基于商标分卡标准g4的分卡数为1,M4值均为1,假设θ的取值为10%,计算结果如下:
Smeaning=[(M1+M2×α1+M3×α2+M4×α3)/M0]-θ=[(1+0+0+1×0)/1]-10%=90%。
例如:输入商标是图2o所示的“格力电器”,假设以输入商标的多种特征类型的分卡集合作为检索关键词对商标存储器进行检索,商标存储器中存储有在先商标的“格力”数据,获取相关的查询结果商标是“格力”,假设θ的取值为10%,按照所述商标义近率的计算公式计算结果商标与输入商标的义近率的过程如下:
输入商标中的“电器”是“商品服务通用名称”,属“例外调整文字”,计算时应予去除;
输入商标“格力电器”与比较的结果商标“格力”基于商标分卡标准g1的分卡数为1,M0和M1值均为1,输入商标“格力电器”与商标分卡标准g2、g3、不适用,M2、M3值均为0,输入商标“格力电器”与比较的结果商标“格力”基于商标分卡标准g4的分卡数为1,M4值为1,θ的取值为10%,计算结果如下:
Smeaning=[(M1+M2×α1+M3×α2+M4×α3)/M0]-θ=[(1+0+0+1×0)/1]-10%=90%。
(4)检索关键词匹配得分率包括如下至少一项:检索关键词匹配综合平均得分率,检索关键词匹配分类平均得分率,检索关键词匹配分类最高得分率,检索关键词匹配分类加权最高得分率,即:Skeywork=S1或Skeywork=S2或Skeywork=S3或Skeywork=S4
其中,Skeywork表示检索关键词匹配得分率,S1表示检索关键词匹配综合平均得分率, S2表示检索关键词匹配分类平均得分率,S3表示检索关键词匹配分类最高得分率,S4表示检索关键词匹配分类加权最高得分率。
各项检索关键词匹配得分率计算方法如下:
检索关键词匹配综合平均得分率S1的计算公式为:S1=(J1+J2+J3+……+Jn)÷n
其中,S1表示检索关键词匹配综合平均得分率,J1、J2、J3……Jn分别表示结果商标与输入商标相匹配的每一分词所对应的商标分卡标准的预设近似度评价分值,n表示结果商标与输入商标相匹配的分卡数。
2)检索关键词匹配分类平均得分率S2的计算公式为:S2=(k1+k2+k3+…+kr)÷r
其中,S2表示检索关键词匹配分类平均得分率,k1表示结果商标与输入商标在第1特征类型相匹配的各分词所对应的商标分卡标准的预设近似度评价分值的平均分值,k2表示结果商标与输入商标在第2特征类型相匹配的各分词所对应的商标分卡标准的预设近似度评价分值的平均分值,k3表示结果商标与输入商标在第3特征类型相匹配的各分词所对应的商标分卡标准的预设近似度评价分值的平均分值,kr表示结果商标与输入商标在第r特征类型相匹配的各分词所对应的商标分卡标准的预设近似度评价分值的平均分值,r表示相匹配的特征类型数。
3)检索关键词匹配分类最高得分率S3的计算公式为:S3=(T1+T2+T3+…+Tr)÷r
其中,S3表示检索关键词匹配分类最高得分率,T1表示结果商标与输入商标在第1特征类型相匹配的各分词所对应的分词标准的预设近似度的评价分值中的最高分值,T2表示结果商标与输入商标在第2特征类型相匹配的各分词所对应的分词标准的预设近似度的评价分值中的最高分值,T3表示结果商标与输入商标在第3特征类型相匹配的各分词所对应的分词标准的预设近似度的评价分值中的最高分值,Tr表示结果商标与输入商标在第r特征类型相匹配的各分词所对应的分词标准的预设近似度的评价分值中的最高分值,r表示相匹配的特征类型数。
4)检索关键词匹配分类加权最高得分率S4的计算公式为:S4=T1×ω1+T2×ω2+T3×ω3+…+Tr×ωr
其中,S4表示检索关键词匹配分类加权最高得分率,T1表示结果商标与输入商标在第1特征类型相匹配的各分词所对应的分词标准的预设近似度评价分值中的最高分值,T2表示结果商标与输入商标在第2特征类型相匹配的各分词所对应的分词标准的预设近似度评价分值中的最高分值,T3表示结果商标与输入商标在第3特征类型相匹配的各分词所对应的分词标准的预设近似度评价分值中的最高分值,Tri表示结果商标与输入商标在第r特征类型相匹配的各分词所对应的分词标准的预设近似度评价分值中的最高分值,r表示相匹配的特征类型数,ω1、ω2、ω3、……ωr分别表示结果商标与输入商标在第1特征类型、第2特征类型、第3特征类型、……、第r特征类型相匹配的各分词所对应的分词标准的预设近似度评价分值中的最高分值的计算权数,ω1、ω2、ω3、……ωr的取值范围在1%至80%之间,全部计算权数的合计为100%。
在一些实施例中,所述特征类型按形音义划分包括:形状特征类型(T1),读音特征类型(T2),含义特征类型(T3);按构成要素内容划分包括:汉语文字特征类型(T1)、字母文字特征类型(T2)、数字文字特征类型(T3)、符号文字特征类型(T4)、图形要素编码图形特征类型(T5)、图像特征描述符图形特征类型(T6)。
例如:输入商标是图2d所示的“四通欧普”,以输入商标的多种特征类型的分卡集合作为检索关键词对商标存储器进行检索,获取相关的查询结果商标是“欧普”和“四通”,检索关键词所匹配的分卡包括根据a11、a12、a13、e1、g1商标分卡标准切分所得到的分卡,并假定a11、a12、a13、e1、g1、j1每一商标分卡标准的预设近似度评价分值分别为50%、60%、40%、40%、100%,形状特征类型(T1)、读音征类型(T2)、含义特征类型(T3)的计算权数分 别为:ω1=50%,ω2=20%,ω3=30%,按照本实施例中检索关键词匹配得分率的计算公式计算结果如下:
1)检索关键词匹配综合平均得分率为:
S1=(J1+J2+J3+……+Jn)÷n=(50%+60%+40%+40%+100%)÷5=58%。
2)检索关键词匹配分类平均得分率为:
按照形音义对商标分词进行划分时,所述特征类型包括形状特征类型、读音特征类型和含义特征类型三种特征类型。则本实施例中,根据a11、a12、a13商标分卡标准得到的分卡属于形状特征类型,根据e1商标分卡标准得到的分词属于读音特征类型,根据g1商标分卡标准得到的分词属于含义特征类型,相匹配的特征类型数r为3。
检索关键词匹配分类平均得分率为:S2=(k1+k2+k3+…+kr)÷r其中,
r=3,
k1=(50%+60%+40%)÷3=50%,
K2=40%÷1=40%,
K3=100%÷1=100%,
因此,S2=(50%+40%+100%)÷3=63.33%。
3)检索关键词匹配分类最高得分率
本实施例中,检索关键词形状特征类型中最高得分的商标分卡标准是a12商标分卡标准,得分为60%,检索关键词读音特征类型中最高得分的商标分卡标准是e1商标分卡标准,得分为40%,检索关键词含义特征类型中最高得分的商标分卡标准是g1商标分卡标准,得分为100%,相匹配的特征类型数r为3。
检索关键词匹配分类最高得分率为:S3=(T1+T2+T3+…+Tr)÷r,其中,
r=3
T1=60%
T2=40%
T3=100%。
因此,S3=(60%+40%+100%)÷3=66.67%。
4)检索关键词匹配分类加权最高得分率
计算公式为:S4=T1×ω1+T2×ω2+T3×ω3+…+Tr×ωr=60%×50%+40%×20%+100%×30%=30%+8%+30%=68%。
第五,步骤S150:按照预设的商标近似度综合量化值的计算公式,计算获取商标近似度综合量化值,然后利用商标近似度综合量化值的大小对结果商标进行排序。
本实施例中,通过以下公式计算商标近似度综合量化值:
TMnear=Wunit×Q1+Ssound×Q2+Smeaning×Q3+Skeywork×Q4
其中,TMnear表示商标近似度综合量化值,Wunit表示商标形近率,Ssound表示商标音近率,Smeaning表示商标义近率,Skeywork表示检索关键词匹配得分率,Q1、Q2、Q3、Q4分别表示商标形近率、商标音近率、商标义近率和检索关键词匹配得分率的权重数,Q1、Q2、Q3、Q4取值范围在5%至95%之间,但全部计算权数的合计为100%。
以下结合一些具体的商标原图实例,对以上商标近似度综合量化值的计算方法进行说明:
假设输入商标是图2o所示的“格力电器”,所获取的结果商标是“格力”、“挌历”,其中,输入商标的“电器”是“商品服务通用名称”,属于商标例外调整文字,经计算获取的检索关键词所匹配的分词包括根据a8、a11、a12、a13、d2、e1、g1分卡标准所切分卡,并设定a8、a11、a12、a13、d2、e1、g1对应的预设近似度评价分值分别为90%、50%、60%、40%、60%、40%、100%,λ1的取值为90%,λ2的取值为80%,μ1的取值为90%,μ2的取值为80%,其中,预 设的商标形近率、商标音近率、商标义近率和检索关键词匹配得分率的权重数取值分别为40%、15%、30%、15%,本实施例按照形音义对商标分词进行划分,所述特征类型包括形状特征类型、读音特征类型和含义特征类型三种特征类型。取检索关键词匹配分类最高得分率作为检索关键词匹配得分率,“电器”是“商品服务通用名称”,属于商标例外调整参数,商标近似度综合量化值的计算过程及结果如下:
1、输入商标“格力电器”与结果商标“格力”的商标
首先,分别计算输入商标“格力电器”与结果商标的商标形近率、商标音近率、商标义近率和检索关键词匹配得分率:
1)商标形近率的计算结果:
Wunit=Ua/(U01)+[Ub/(U01)]×λ1–[Uc/(U01)]×λ2=2/(2-0)+0/(2-0)×90%–0/(2-0)×80%=100%。
2)商标音近率的计算结果:
输入商标“格力电器”的读音为“ge”、“li”、“dian”、“qi”,结果商标“格力”的读音为“ge”、“li”
Ssound=Va/(V02)+[Vb/(V02)]×μ1–[Vc/(V02)]×μ2=2/(2-0)+0/(2-0)×90%–0/(2-0)×80%=100%。
3)商标义近率的计算结果:
“电器”属例外调整文字,输入商标“格力电器”去除例外调整文字后是“格力”,输入商标“格力电器”去除例外调整文字后的“格力”与比较的结果商标“格力”相匹配,属输入商标去除例外调整文字后与结果商标基于商标分卡标准g1相匹配的分卡,M0、M1均为1。本实施行中,M2、M3均为0,“格力”在汉语词典中没有记载,属无含义的组合,故M4为1。输入商标与结果商标文字数不相同,符合θ的调整参数特征,θ为10%,则:
Smeaning=[(M1+M2×α1+M3×α2+M4×α3)/M0]-θ=[(1+0+0+1×0)/1]-10%=90%。
4)检索关键词匹配得分率:本实施例取检索关键词匹配分类最高得分率的计算过程如下:
检索关键词形状特征类型中最高得分T1是商标分卡标准a8的商标分卡标准,得分为90%,检索关键词读音特征类型中最高得分T2是商标分卡标准e1的商标分卡标准,得分为40%,检索关键词含义特征类型中最高得分T3是商标分卡标准g1的商标分卡标准,得分为100%,相匹配的特征类型数r为3。
因此,Skeywork=(T1+T2+T3+…+Tr)÷r=(90%+40%+100%)÷3=76.67%。
然后,根据输入商标“格力电器”与结果商标的商标形近率、商标音近率、商标义近率和检索关键词匹配得分率计算其商标近似度综合量化值:
TMnear=Wunit×Q1+Ssound×Q2+Smeaning×Q3+Skeywork×Q4
=100%×40%+100%×15%+90%×30%+76.67%×15%
=40%+15%+27%+11.5%=93.5%。
2、输入商标“格力电器”与结果商标“挌历”的商标
首先,分别计算输入商标“格力电器”与结果商标的商标形近率、商标音近率、商标义近率和检索关键词匹配得分率:
1)商标形近率的计算结果:
Wunit=Ua/(U01)+[Ub/(U01)]×λ1–[Uc/(U01)]×λ2=0/(2-0)+2/(2-0)×90%–0/(2-0)×80%=90%。
2)商标音近率的计算结果:
输入商标“格力电器”的读音为“ge”、“li”、“dian”、“qi”,与结果商标“挌历”的读音为“ge”、“li”
Ssound=Va/(V02)+[Vb/(V02)]×μ1–[Vc/(V02)]×μ2=0/(2-0)+[2/(2-0)]×90%–[0/(2-0)]×80%=90%。
3)商标义近率的计算结果:
“电器”属例外调整文字,输入商标“格力电器”去除例外调整文字后是“格力”,输入商标“格力电器”去除例外调整文字后的“格力”与比较的结果商标“挌历”相匹配,属输入商标去除例外调整文字后与结果商标的商标基于分卡标准g2相匹配的分卡,M0、M2均为1,M1、M3分卡数均为0,“格力”在汉语词典中没有记载,属无含义的组合,故M4为1。输入商标与结果商标文字数不相同,符合θ的调整参数特征,θ为10%,则:
Smeaning=[(M1+M2×α1+M3×α2+M4×α3)/M0]-θ=[(0+1×1+0+1×0)/1]-10%=90%。
4)检索关键词匹配得分率,本实施例取检索关键词匹配分类最高得分率的计算过程如下:
检索关键词形状特征类型中最高得分T1是商标分卡标准a8的商标分卡标准,得分为90%,检索关键词读音特征类型中最高得分T2是商标分卡标准e1的商标分卡标准,得分为40%,检索关键词含义特征类型中最高得分T3是商标分卡标准g1的商标分卡标准,得分为100%,相匹配的特征类型数r为3。
因此,Skeywork=(T1+T2+T3+…+Tr)÷r=(90%+40%+100%)÷3=76.67%。
然后,根据输入商标“格力电器”与结果商标的商标形近率、商标音近率、商标义近率和检索关键词匹配得分率计算其商标近似度综合量化值:
TMnear=Wunit×Q1+Ssound×Q2+Smeaning×Q3+Skeywork×Q4
=100%×40%+100%×15%+100%×30%+76.67%×15%
=40%+15%+30%+11.5%=96.5%。
最后利用商标近似度综合量化值的大小对结果商标进行排序,即可清晰展示进一步符合《商标法》意义上的商标相同或近似的要求的结果商标检索列表。
图5示出了采用商标近似度综合量化值排序的前24件结果商标报告界面截图。该实施例以图2n所示的图形为输入商标,商品范围是尼斯分类的第42类,注册国是中国,结果商标经本发明前述方法的商标近似度综合量化值计算获取后所得前24件商标报告界面截图。
本发明所述的商标查询结果近似度评价和排序方法,可以有效克服传统的商标查询结果单一特征排序方法造成的排序结果片面或漏检的缺陷和弊端,能全面反映商标的形、音、义相结合的综合特征,提升商标相同或近似判断的准确性和查全率。采用商标近似度综合量化值,有效地将商标图像抽象的视觉结果进行量化,大幅提高商标近似度的数量化评价水平。本发明提高了商标相同或近似判断的标准化水平,缩小了商标查询结果近似度排序结果与审查人员预期的《商标法》意义上的商标相同或近似的排序结果的差异,实现输入商标与样本商标是否构成商标相同或近似的较好评价,加速商标审查工作的进步。本发明只需将待检索的商标一次性输入系统即可得到最佳的综合排序结果,克服现有商标检索系统需要不断地进行人机交互以获取不同的排序和显示结果,或利用人工筛选而造成的检索结果主观性太强的缺陷。
在本发明的实施例中,还涉及一种商标查询结果近似度评价和排序的装置,图6是本发明实施例中的商标查询结果近似度评价和排序装置的结构示意图,一种商标查询结果近似度评价和排序装置包括:
样本商标分卡预处理模块:用于对样本商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对样本商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容,(3)、样本商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为样本商标分卡信息,并设定每一预设的商标分卡标准的近似度评价分值;
输入商标分卡处理模块:用于对输入商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对输入商标是否由汉语文字,图形,字母、数字或符号构成要素进行识别,获取构成要素的内容;(3)、提取输入商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为输入商标分卡信息;
商标检索模块:以输入商标分卡信息集合作为检索关键词对存储于商标存储器的样本商标分卡信息进行检索,获取相关的结果商标的分卡信息及分卡匹配信息;
商标形近率计算模块:用于按照预设的商标形近率计算公式,计算输入商标与结果商标之间的商标形近率;
商标义近率计算模块:用于按照预设的商标义近率计算公式,计算输入商标与结果商标之间的商标义近率;
商标音近率计算模块:用于按照预设的商标音近率计算公式,计算输入商标与结果商标之间的商标音近率;
检索关键词匹配得分率计算模块:用于按照预设的检索关键词匹配得分率计算公式,计算输入商标与结果商标之间的检索关键词匹配得分率;
商标近似度综合量化值的计算模块:用于按照预设的商标近似度综合量化值的计算公式,计算获取商标近似度综合量化值,然后利用商标近似度综合量化值的大小对结果商标进行排序。
实施例二:
本实施例提供一种商标查询结果近似度评价和排序方法,与实施例一的区别仅在于:所述商标查询结果近似度评价和排序方法中的前两个步骤的顺序不同,本实施例具体包括以下步骤:
步骤S210:对输入商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对输入商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容;(3)、提取输入商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为输入商标分卡信息;
步骤S220:对样本商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对样本商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容,(3)、样本商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为样本商标分卡信息,并设定每一预设的商标分卡标准的近似度评价分值;
步骤S230:以输入商标分卡信息集合作为检索关键词对存储于商标存储器的样本商标分卡信息进行检索,获取相关的结果商标的分卡信息及分卡匹配信息;
步骤S240:按照预设的商标形近率、商标义近率、商标音近率和检索关键词匹配得分率计算公式,分别计算输入商标与结果商标之间的商标形近率、商标义近率、商标音近率和检索关键词匹配得分率;
步骤S250:按照预设的商标近似度综合量化值的计算公式,计算获取商标近似度综合 量化值,然后利用商标近似度综合量化值的大小对结果商标进行排序。
以上,结合具体的实施例对本发明的技术方案进行了详细介绍,所描述的具体实施例用于帮助理解本发明的思想,但并不能因此而理解为对本发明保护范围的限制。应当指出,本领域技术人员在本发明实施例的基础上做出的变形、推导、变换,也都应属于本发明保护范围之内。

Claims (8)

  1. 一种商标查询结果近似度评价和排序方法,其特征在于,包括以下步骤:
    步骤S110:对样本商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对样本商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容,(3)、样本商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为样本商标分卡信息,并设定每一预设的商标分卡标准的近似度评价分值;
    步骤S120:对输入商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对输入商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容;(3)、提取输入商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为输入商标分卡信息;
    步骤S130:以输入商标分卡信息集合作为检索关键词对存储于商标存储器的样本商标分卡信息进行检索,获取相关的结果商标的分卡信息及分卡匹配信息;
    步骤S140:按照预设的商标形近率、商标义近率、商标音近率和检索关键词匹配得分率计算公式,分别计算输入商标与结果商标之间的商标形近率、商标义近率、商标音近率和检索关键词匹配得分率;
    步骤S150:按照预设的商标近似度综合量化值的计算公式,计算获取商标近似度综合量化值,然后利用商标近似度综合量化值的大小对结果商标进行排序。
  2. 根据权利要求1所述商标查询结果近似度评价和排序方法,其特征在于,步骤S110和步骤S120所述“形状特征最小单元、读音特征最小单元、含义特征最小单元”和“商标分卡标准”包括:
    1)形状特征最小单元包括:
    构成要素为汉语文字的形状特征最小单元,选如下之一:每一汉语文字,每一汉语文字的每一笔划;
    构成要素为图形的形状特征最小单元,选如下之一:商标图形要素编码,预设定长的商标图像轮廓线上像素点集;
    构成要素为字母的形状特征最小单元,选如下之一:每一语种的单词,每一个字母;
    构成要素为汉语数字的形状特征最小单元,选如下之一:汉语数字的组合,每一单个的汉语数字;
    构成要素为阿拉伯数字的形状特征最小单元,选如下之一:阿拉伯数字的组合,每一单个的阿拉伯数字;
    构成要素为其他语种数字的形状特征最小单元,选如下之一:其他语种数字的组合,每一单个的其他语种数字;
    构成要素为符号的形状特征最小单元:为每一单个的符号;
    2)含义特征最小单元包括:
    构成要素为汉语文字的含义特征最小单元:商标的整体汉语文字组合由汉语词典所记载的词语组合构成时,每个词语为含义特征最小单元,否则,商标的整体汉语文字组合为含义特征最小单元;
    构成要素为图形的含义特征最小单元:商标图形要素编码所对应的每一事物的名称;
    构成要素为字母的含义特征最小单元:商标的整体字母组合由英文词典所记载的单词组合构成,或其他语种词典所记载的单词组合构成时,每个单词为含义特征最小单元,否则,商标的整体字母组合为含义特征最小单元;
    构成要素为汉语数字的含义特征最小单元,选如下之一:商标中被分隔的每一组汉语数字所对应的预设基准语种数字,商标中每一单个汉语数字所对应的预设基准语种数字,其中,所述预设基准语种数字为任何语种数字;
    构成要素为阿拉伯数字的含义特征最小单元,选如下之一:商标中被分隔的每一组阿拉伯数字所对应的预设基准语种数字,商标中每一单个阿拉伯数字所对应的预设基准语种数字,其中,所述预设基准语种数字为任何语种数字;
    构成要素为其他语种数字的含义特征最小单元,选如下之一:商标中被分隔 的每一组其他语种数字所述对应的预设基准语种数字,商标中每一单个的其它语种数字所对应的预设基准语种数字,所述预设基准语种数字为任何语种数字;
    构成要素为符号的含义特征最小单元:商标中每一符号所对应的符号名称;
    3)读音特征最小单元包括:
    构成要素为汉语文字的读音特征最小单元,为每一汉语文字的拼音;
    构成要素为图形的读音特征最小单元,为商标图形要素编码所对应的每一事物的名称的拼音;
    构成要素为字母的读音特征最小单元,选如下之一:每一字母组合的读音、每一个字母的读音;
    构成要素为数字或符号的读音特征最小单元,选如下之一:商标中被分隔的每一组数字的读音,每一单个数字的读音,商标中被分隔的每一组符号的读音,每一单个符号的读音;
    4)商标分卡标准包括:
    A、由构成要素为汉语文字的形状特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准a1、a2、a3、a4、a5、a6、a7、a8、a9、a10、a11、a12、a13中的至少一种,其中,
    a1表示按顺序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡,
    a2表示按逆序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡,
    a3表示按顺序排列的商标中的汉语文字切分为一个分卡,
    a4表示按逆序排列的商标中的汉语文字切分为一个分卡,
    a5表示按顺序排列的商标中含有汉语数字切分为一个分卡,
    a6表示按逆序排列的商标中含有汉语数字切分为一个分卡,
    a7表示商标中每个相对独立的部分分别切分为一个分卡,
    a8表示商标文字中完整包含现有汉语文字商标,将该部分切分为一个分卡,
    a9表示商标中含有的繁体、异体字转换为简体字后切分为一个分卡,
    a10表示商标中每个文字替换为形近字后切分为一个分卡,
    a11表示商标中每相邻汉语文字分别切分为一个分卡,
    a12表示商标中首尾汉字组合切分为一个分卡,
    a13表示商标中每个汉语文字切分为一个分卡;
    B、由构成要素为字母、数字、符号的形状特征最小单元多种组合方案所构成的商标分卡标准,,包括:分卡标准b1、b2、b3、b4、b5、b6、b7、b8、b9、b10、b11、b12、b13、b14中的至少一种,其中,
    b1表示按顺序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡,
    b2表示按逆序排列的商标整体所有语种文字及图形要素编码的组合切分为一个分卡,
    b3表示按顺序排列的商标中的字母组合切分为一个分卡,
    b4表示按逆序排列的商标中的字母组合切分为一个分卡,
    b5表示按顺序排列的商标中含有的非汉语数字或每一单个的非汉语数字分别切分为一个分卡,
    b6表示按逆序排列的商标中含有的非汉语数字或每一单个的非汉语数字分别切分为一个分卡,
    b7表示按顺序排列的商标中含有的符号组合切分为一个分卡,
    b8表示按逆序排列的商标中含有的符号组合切分为一个分卡,
    b9表示商标中每个相对独立部分分别切分为一个分卡,
    b10表示商标中每字母被形近字母替换后切分为一个分卡,
    b11表示商标中每相邻字母组合分别切分为一个分卡,
    b12表示商标中字母按照不同定序排列后分别切分为一个分卡,
    b13表示商标中首尾字母组合切分为一个分卡,
    b14表示商标中每个字母或数字或符号分别切分为一个分卡;
    C、由构成要素为图形的形状特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准c1、c2、c3、c4中的至少一种,其中,
    c1表示商标图形要素编码集合整体切分为一个分卡,
    c2表示每一商标图形要素编码切分为一个分卡,
    c3表示每种图像特征识别方法所产生的商标图像特征描述符整体分别切分为一个分卡,
    c4表示每种图像特征识别方法所产生的商标图像特征描述符预设定长分别切分为一个分卡,所述商标图像特征描述符预设定长是指预先设置的商标图像轮廓线上连续相连的像素点的长度,连续相连的像素点以特征字符串集、或数字集表示,取值范围为商标图像特征描述符、或数字集总长度的0.1%-50%;
    D、由构成要素为汉语文字的读音特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准d1、d2、d3中的至少一种,其中,
    d1表示商标中每个汉语文字的拼音音节切分为一个分卡,
    d2表示商标中整体汉语文字对应的拼音切分为一个分卡,
    d3表示商标中每个中文文字被替换为形近字后的拼音音节切分为一个分卡;
    E、由构成要素为字母、数字、符号的读音特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准e1、e2、e3、e4中的至少一种,其中,
    e1表示商标中每个英文单词的读音音节切分为一个分卡,
    e2表示商标中字母组合被音近字母组合替换后得到的整体字母组合分别切分为一个分卡,
    e3表示商标中每个数字的读音音节切分为一个分卡,
    e4表示商标中每个符号的读音音节切分为一个分卡;
    F、由构成要素为图形的读音特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准f1,其中,f1表示商标的图形要素编码所对应的每一事物的名称的拼音切分为一个分卡;
    G、由构成要素为汉语文字的含义特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准g1、g2、g3、g4中的至少一种,其中,
    g1表示商标中完整包含商标服务器中现有汉语文字商标、且商标整体无含义,将含有现有汉语文字商标的部分切分为一个分卡,
    g2表示商标中与汉语词典所收录的词汇或商标服务器中现有汉语文字商标的汉字组合全部匹配的,将匹配部分分别切分为一个分卡,
    g3表示商标中含有的汉语词汇替换为近义词后分别切分为一个分卡,
    g4表示商标整体无含义的,整体汉语文字切分为一个分卡;
    H、由构成要素为字母、数字、符号组合的含义特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准h1、h2、h3、h4、h5、h6、h7、h8、h9 中的至少一种,其中,
    h1表示商标的整体字母组合由英文词典或其他语种词典所记载的单词组合构成,整体单词组合切分为一个分卡,
    h2表示商标中含有英语词典或其他语种词典所收录的单词,每个单词切分为一个分卡,
    h3表示商标中含有英语词典或其他语种词典所收录的单词,将每个单词的近义词切分为一个分卡,
    h4表示商标的整体字母组合与英文词典或其他语种词典所记载的单词不匹配的,整体字母组合切分为一个分卡,
    h5表示商标中被分隔的每一组数字切分为一个分卡,
    h6表示商标的整体数字组合切分为一个分卡,
    h7表示商标的整体符号组合切分为一个分卡,
    h8表示商标的每一符号切分为一个分卡,
    h9表示商标完整包含商标服务器现有字母组合商标、且商标整体无含义,将含有现有字母组合商标的部分切分为一个分卡;
    I、由构成要素为图形的含义特征最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准i1、i2中的至少一种,其中,
    i1表示商标图形要素编码所对应的每一事物的名称切分为一个分卡,
    i2表示商标图像特征描述符对应有商标图形要素编码,商标图形要素编码所对应每个事物名称切分为一个分卡;
    Y、由构成要素为例外调整文字的最小单元多种组合方案所构成的商标分卡标准,包括:分卡标准y1、y2中的至少一种,其中,
    y1表示商标中含有例外调整文字,将例外调整文字整体切分为一个分卡;
    y2表示商标中含有例外调整文字,将例外调整文字每个文字分别切分为一个分卡。
  3. 根据权利要求2所述商标查询结果近似度评价和排序方法,其特征在于,其中,所述例外调整文字包括如下的一种以上的文字:县级以上行政区域地名,公众知晓的外国地名,通用商品名称,表示商品的质量、主要原料、功能、用途、重量、数量及其他特点的词语,商品服务通用名称,显著性弱的文字。
  4. 根据权利要求2所述商标查询结果近似度评价和排序方法,其特征在于,步骤S120所述“输入商标分卡信息”包括:U0、β1、V0、β2、M0、Y0,其中,U0表示输入商标基于商标分卡标准a13、b14、c2、c4或其组合所得的分卡数;β1表示输入商标中含有的例外调整文字的基于分卡标准a13、b14、c2、c4所得分卡数或文字数;V0表示输入商标基于商标分卡标准d1、d2、d3、e1、e2、e3、e4或其组合所得的分卡数;β2表示输入商标中含有的例外调整文字的基于分卡标准d1、d2、d3、e1、e2、e3、e4所得分卡数或音节数;M0表示输入商标去除例外调整文字后与结果商标基于商标分卡标准g1、g2、g3、g4所得分卡中相匹配的分卡数;Y0表示输入商标基于商标分卡标准y1或y2所得的分卡数;
    步骤S130所述“结果商标的分卡信息及分卡匹配信息”包括Ya、Ua、Ub、Uc、Va、Vb、Vc、M1、M2、M3、M4、Ji、n、ki、r、Ti,其中,Ya表示结果商标基于商标分卡标准y1或y2所得的分卡数;Ua表示结果商标去除例外调整文字后与输入商标基于商标分卡标准a13、b14、c2、c4或其组合所得分卡中相匹配的分卡数;Ub表示结果商标去除例外调整文字后与输入商标基于商标分卡标准a10、b10或其组合所得分卡中相匹配的分卡数;Uc表示结果商标与输入商标基于商标分卡标准a13、b14、c2、c4或其组合和a10、b10或其组合所得相匹配的分卡中间插入不匹配分卡的处数;Va表示结果商标去除例外调整文字后与输入商标基于商标分卡标准d1、d2、e1、e3、e4或其组合所得分卡中相匹配的分卡数;Vb表示结果商标去除例外调整文字后与输入商标基于商标分卡标准d3、e2或其组合所得分卡中相匹配的分卡数;Vc表示结果商标与输入商标基于商标分卡标准d1、d2、e1、e3、e4或其组合和d3、e2或其组合所得相匹配的分卡中间插入不匹配分卡的处数;M1表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g1的匹配分卡数;M2表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g2的匹配分卡数;M3表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g3的匹配分卡数,M4表示比较的结果商标去除例外调整文字后与输入商标基于商标分卡标准g4的匹配分卡数;Ji表示结果商标与输入商标相匹配的第i个分卡所对应的商标分卡标准的预设近似度评价分值;n表示结果商标与输入商标相匹配的分卡项目数;ki表示结果商标与输入商标在第i特征类型相匹配的各分词所对应的商标分卡标准的预设近似度评价分值的平均分值, r表示结果商标与输入商标相匹配的特征类型数;Ti表示结果商标与输入商标在第i特征类型相匹配的各分词所对应的分词标准的预设近似度的评价分值中的最高分值;
    所述特征类型通过预设的分类标准对商标分卡信息进行分类所得到的分卡类别。
  5. 根据权利要求4所述商标查询结果近似度评价和排序方法,其特征在于,其中,所述特征类型,按形音义划分包括:形状特征类型、读音特征类型、含义特征类型;按构成要素内容划分包括:汉语文字特征类型、字母文字特征类型、数字文字特征类型、符号文字特征类型、图形要素编码图形特征类型、图像特征描述符图形特征类型。
  6. 根据权利要求4所述商标查询结果近似度评价和排序方法,其特征在于,其中,步骤S140所述“预设的商标形近率、商标义近率、商标音近率和检索关键词匹配得分率计算公式”,包括:
    1)商标形近率计算公式包括:
    Wunit=Ua/(U01)+[Ub/(U01)]×λ1–[Uc/(U01)]×λ2
    其中,Wunit表示商标形近率,λ1、λ2为预设的调整权数,λ1和λ2的取值范围均在10%至300%之间;
    2)商标音近率计算公式包括:
    Ssound=Va/(V02)+[Vb/(V02)]×μ1–[Vc/(V02)]×μ2
    其中,Ssound表示商标音近率,μ1、μ2为预设的调整权数,取值范围均在10%至300%之间;
    3)商标义近率计算公式包括:
    Smeaning=[(M1+M2×α1+M3×α2+M4×α3)/M0]-θ
    其中,Smeaning表示商标义近率,α1、α2、α3分别表示对M2、M3、M4的调整参数,取值规则:当出现M1、M2、M3、M4中二个及以上参数同时不为0时,在M1、M2、M3、M4中排列最前参数为有效参数,其余为无效参数,当M1不为0时,α1、α2、α3的值为0;当M1为0且M2不为0时,α1为1,α2、α3的值为0;当M1、M2为0且M3不为0时,α2为1,α3为0;当M1、M2、M3为0且M4不为0时,α3为1;θ表示输入商标与比较的结果商标文字数不 相同的调整参数,取值范围在1%至90%之间;
    4)检索关键词匹配得分率计算公式包括如下至少一项:检索关键词匹配综合平均得分率,检索关键词匹配分类平均得分率,检索关键词匹配分类最高得分率,检索关键词匹配分类加权最高得分率,即:
    Skeywork=S1或Skeywork=S2或Skeywork=S3或Skeywork=S4
    其中,Skeywork表示检索关键词匹配得分率,S1表示检索关键词匹配综合平均得分率,S2表示检索关键词匹配分类平均得分率,S3表示检索关键词匹配分类最高得分率,S4表示检索关键词匹配分类加权最高得分率;
    其计算公式分别为:
    S1=(J1+J2+J3+……+Jn)÷n
    S2=(k1+k2+k3+…+kr)÷r
    S3=(T1+T2+T3+…+Tr)÷r
    S4=T1×ω1+T2×ω2+T3×ω3+…+Tr×ωr
    其中,ω1、ω2、ω3、……ωr分别表示结果商标与输入商标在第1特征类型、第2特征类型、第3特征类型、……、第r特征类型相匹配的各分卡所对应的分卡标准的预设近似度评价分值中的最高分值的计算权数,ω1、ω2、ω3、……ωr的取值范围在1%至80%之间,全部计算权数的合计为100%。
  7. 根据权利要求6所述商标查询结果近似度评价和排序方法,其特征在于,其中,步骤S150所述“商标近似度综合量化值的计算公式”包括:
    TMnear=Wunit×Q1+Ssound×Q2+Smeaning×Q3+Skeywork×Q4
    其中,TMnear表示商标近似度综合量化值,Wunit表示商标形近率,Ssound表示商标音近率,Smeaning表示商标义近率,Skeywork表示检索关键词匹配得分率,Q1、Q2、Q3、Q4分别表示商标形近率、商标音近率、商标义近率和检索关键词匹配得分率的权重数,Q1、Q2、Q3、Q4取值范围在5%至95%之间,全部计算权数的合计为100%。
  8. 一种商标查询结果近似度评价和排序装置,其特征在于,包括:
    样本商标分卡预处理模块:用于对样本商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对样 本商标是否由汉语文字、图形、字母、数字或符号构成要素进行识别,获取构成要素的内容,(3)、样本商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为样本商标分卡信息,并设定每一预设的商标分卡标准的近似度评价分值;
    输入商标分卡处理模块:用于对输入商标图像及内容按预设的商标分卡标准进行商标分卡处理,具体处理过程包括:(1)、建立由预设的形状特征、读音特征和含义特征最小单元多种组合方案所构成的商标分卡标准,(2)、对输入商标是否由汉语文字,图形,字母、数字或符号构成要素进行识别,获取构成要素的内容;(3)、提取输入商标各构成要素的形状特征最小单元、读音特征最小单元和含义特征最小单元;(4)、根据已建立的商标分卡标准,提取每一组合方案所生成或转换得到的各种文字、图形的切分信息,将这些切分信息作为输入商标分卡信息;
    商标检索模块:以输入商标分卡信息集合作为检索关键词对存储于商标存储器的样本商标分卡信息进行检索,获取相关的结果商标的分卡信息及分卡匹配信息;
    商标形近率计算模块:用于按照预设的商标形近率计算公式,计算输入商标与结果商标之间的商标形近率;
    商标义近率计算模块:用于按照预设的商标义近率计算公式,计算输入商标与结果商标之间的商标义近率;
    商标音近率计算模块:用于按照预设的商标音近率计算公式,计算输入商标与结果商标之间的商标音近率;
    检索关键词匹配得分率计算模块:用于按照预设的检索关键词匹配得分率计算公式,计算输入商标与结果商标之间的检索关键词匹配得分率;
    商标近似度综合量化值的计算模块:用于按照预设的商标近似度综合量化值的计算公式,计算获取商标近似度综合量化值,然后利用商标近似度综合量化值的大小对结果商标进行排序。
PCT/CN2017/100187 2016-12-30 2017-09-01 一种商标查询结果近似度评价和排序方法、装置 WO2018120899A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/475,333 US20200387543A1 (en) 2016-12-30 2017-09-01 Trademark inquiry result proximity evaluating and sorting method and device

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN201611257312.6 2016-12-30
CN201611257312.6A CN106649851A (zh) 2016-12-30 2016-12-30 近似商标查询结果排序方法、装置及其商标服务器
CN201710553009.9 2017-07-07
CN201710553009.9A CN107330109B (zh) 2016-12-30 2017-07-07 一种商标查询结果近似度评价和排序方法、装置

Publications (1)

Publication Number Publication Date
WO2018120899A1 true WO2018120899A1 (zh) 2018-07-05

Family

ID=58837704

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/100187 WO2018120899A1 (zh) 2016-12-30 2017-09-01 一种商标查询结果近似度评价和排序方法、装置

Country Status (3)

Country Link
US (1) US20200387543A1 (zh)
CN (5) CN106649851A (zh)
WO (1) WO2018120899A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113377906A (zh) * 2021-06-08 2021-09-10 四川大学 一种相似法条智能搜索系统及方法
CN113780268A (zh) * 2020-12-28 2021-12-10 北京沃东天骏信息技术有限公司 商标识别方法、装置与电子设备

Families Citing this family (31)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106649851A (zh) * 2016-12-30 2017-05-10 徐庆 近似商标查询结果排序方法、装置及其商标服务器
CN107273535A (zh) * 2017-06-29 2017-10-20 朱峰 一种商标智能分析系统
WO2019028618A1 (zh) * 2017-08-07 2019-02-14 深圳益强信息科技有限公司 基于大数据的商标价值评估的方法及系统
CN107609057B (zh) * 2017-08-25 2020-12-22 百度在线网络技术(北京)有限公司 一种获取商标图像的文字数据的方法与装置
CN108664945B (zh) * 2018-05-18 2021-08-10 徐庆 图像文本及形音义特征识别方法和装置
CN108763380B (zh) * 2018-05-18 2022-03-08 徐庆 商标识别检索方法、装置、计算机设备和存储介质
CN110580666A (zh) * 2018-06-08 2019-12-17 成都市卓睿科技有限公司 商标监控与预警方法及系统
CN110647639B (zh) * 2018-06-08 2020-11-10 成都市卓睿科技有限公司 一种商标近似检索结果排序方法
CN108897722A (zh) * 2018-06-26 2018-11-27 重庆智荟数创科技有限公司 基于笔顺算法的商标近似评估、监控系统及方法
CN108984649A (zh) * 2018-06-27 2018-12-11 广州朝舜网络科技有限公司 一种近似商标智能判断方法、装置、终端及存储介质
CN109033370A (zh) * 2018-07-27 2018-12-18 阿里巴巴集团控股有限公司 一种查找相似店铺的方法及装置、店铺接入的方法及装置
CN109063197B (zh) * 2018-09-06 2021-07-02 徐庆 图像检索方法、装置、计算机设备和存储介质
CN110895555B (zh) * 2018-09-13 2022-06-14 北京蓝灯鱼智能科技有限公司 数据检索方法和装置、存储介质及电子装置
CN109345454B (zh) * 2018-09-18 2023-01-06 徐庆 位图图像矢量化的方法、存储介质及系统
CN109656954A (zh) * 2018-11-28 2019-04-19 苏州中知联信息科技有限公司 商标查询方法、装置及计算机设备
CN109800340B (zh) * 2019-01-24 2021-03-19 北京梦知网科技有限公司 商标注册推荐方法及系统
CN110059159A (zh) * 2019-04-15 2019-07-26 重庆天蓬网络有限公司 一种近似商标实时监控系统
CN110069554A (zh) * 2019-05-06 2019-07-30 重庆天蓬网络有限公司 一种基于商标注册信息的处理方法
CN110069555A (zh) * 2019-05-06 2019-07-30 重庆天蓬网络有限公司 一种商标注册成功率的确定方法
CN110288264A (zh) * 2019-07-03 2019-09-27 深圳智高点知识产权运营有限公司 一种企业商标监控分析系统和方法
CN110717874B (zh) * 2019-10-10 2022-11-04 徐庆 一种图像轮廓线平滑处理方法
CN110929084B (zh) * 2019-12-17 2023-04-11 徐庆 一种图像形状特征描述符的获取方法及装置
CN111125160A (zh) * 2019-12-26 2020-05-08 广东知得失网络科技有限公司 一种基于商标近似分析的数据预处理方法、系统及终端
CN111782851A (zh) * 2020-05-19 2020-10-16 知昇(上海)人工智能科技有限公司 一种基于多相似度特征计算的中文商标名称相似判别方法
CN111882462B (zh) * 2020-08-03 2023-05-09 安徽大学 一种面向多要素审查标准的中文商标近似检测方法
CN113553463B (zh) * 2021-07-30 2024-06-25 徐庆 一种商标识别查询方法、系统、数据存储器及存储介质
CN113553980B (zh) * 2021-07-30 2024-07-19 徐庆 一种图片的商标图形要素编码生成方法、系统及装置
CN113554639A (zh) * 2021-07-30 2021-10-26 徐庆 一种图像特征提取及描述符获取方法、装置及存储介质
CN114581196B (zh) * 2022-03-10 2024-10-18 广州华多网络科技有限公司 商品排序处理方法及其装置、设备、介质、产品
CN115100665B (zh) * 2022-07-22 2024-10-01 贵州中烟工业有限责任公司 近似商标筛选方法、模型构建方法及计算机可读存储介质
CN115774548B (zh) * 2023-02-10 2023-04-11 北京一平方科技有限公司 基于人工智能的代码自动生成方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105426530A (zh) * 2015-12-15 2016-03-23 徐庆 一种商标检索方法、装置和系统
CN105574161A (zh) * 2015-12-15 2016-05-11 徐庆 一种商标图形要素识别方法、装置和系统
CN106649851A (zh) * 2016-12-30 2017-05-10 徐庆 近似商标查询结果排序方法、装置及其商标服务器

Family Cites Families (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5548507A (en) * 1994-03-14 1996-08-20 International Business Machines Corporation Language identification process using coded language words
US5790126A (en) * 1995-01-03 1998-08-04 Microsoft Corporation Method for rendering a spline for scan conversion of a glyph
US5864639A (en) * 1995-03-27 1999-01-26 Digital Processing Systems, Inc. Method and apparatus of rendering a video image
CN101013420A (zh) * 2006-12-31 2007-08-08 中国科学院计算技术研究所 一种识别中文文本编码形式的方法
CN101551859B (zh) * 2008-03-31 2012-01-04 夏普株式会社 图像辨别装置及图像检索装置
CN101567048B (zh) * 2008-04-21 2012-06-06 夏普株式会社 图像辨别装置及图像检索装置
CN101359367B (zh) * 2008-09-11 2010-09-29 西安理工大学 一种基于四边形骨架结构的静态手势特征描述方法
US20120144499A1 (en) * 2010-12-02 2012-06-07 Sky Castle Global Limited System to inform about trademarks similar to provided input
CN103020596B (zh) * 2012-12-05 2016-06-22 华北电力大学 一种基于块模型的电力生产中异常人体行为识别方法
CN103258037A (zh) * 2013-05-16 2013-08-21 西安工业大学 一种针对多组合内容的商标识别检索方法
CN104809142B (zh) * 2014-01-29 2018-03-23 北京瑞天科技有限公司 商标查询系统和方法
CN104462380A (zh) * 2014-12-11 2015-03-25 北京中细软移动互联科技有限公司 商标检索方法
CN105574533B (zh) * 2015-12-15 2018-01-12 徐庆 一种图像特征提取方法和装置
CN105740872B (zh) * 2016-01-29 2020-05-19 徐庆 图像特征提取方法及其装置
CN105809138B (zh) * 2016-03-15 2019-01-04 武汉大学 一种基于分块识别的道路警告标志检测与识别方法
CN105913067A (zh) * 2016-04-18 2016-08-31 徐庆 图像轮廓特征的提取方法及其装置
CN106295656B (zh) * 2016-08-03 2017-09-15 徐庆 基于图像色块内容的图像轮廓特征提取方法和装置

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105426530A (zh) * 2015-12-15 2016-03-23 徐庆 一种商标检索方法、装置和系统
CN105574161A (zh) * 2015-12-15 2016-05-11 徐庆 一种商标图形要素识别方法、装置和系统
CN106649851A (zh) * 2016-12-30 2017-05-10 徐庆 近似商标查询结果排序方法、装置及其商标服务器

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113780268A (zh) * 2020-12-28 2021-12-10 北京沃东天骏信息技术有限公司 商标识别方法、装置与电子设备
CN113377906A (zh) * 2021-06-08 2021-09-10 四川大学 一种相似法条智能搜索系统及方法

Also Published As

Publication number Publication date
CN108052653A (zh) 2018-05-18
CN107330438A (zh) 2017-11-07
CN107330109A (zh) 2017-11-07
CN107301244B (zh) 2018-06-15
US20200387543A1 (en) 2020-12-10
CN107301244A (zh) 2017-10-27
CN107330109B (zh) 2018-04-17
CN107330438B (zh) 2018-04-17
CN106649851A (zh) 2017-05-10

Similar Documents

Publication Publication Date Title
WO2018120899A1 (zh) 一种商标查询结果近似度评价和排序方法、装置
CN108717406B (zh) 文本情绪分析方法、装置及存储介质
CN106156204B (zh) 文本标签的提取方法和装置
CN109189901B (zh) 一种智能客服系统中自动发现新分类以及对应语料的方法
CN110297988B (zh) 基于加权LDA和改进Single-Pass聚类算法的热点话题检测方法
El et al. Authorship analysis studies: A survey
CN103198057B (zh) 一种自动给文档添加标签的方法和装置
CN103678576B (zh) 基于动态语义分析的全文检索系统
CN110222160A (zh) 智能语义文档推荐方法、装置及计算机可读存储介质
CN106599054B (zh) 一种题目分类及推送的方法及系统
CN106651696B (zh) 一种近似题推送方法及系统
CN111611356B (zh) 信息查找方法、装置、电子设备及可读存储介质
CN112667794A (zh) 一种基于孪生网络bert模型的智能问答匹配方法及系统
CN110134792B (zh) 文本识别方法、装置、电子设备以及存储介质
US20130036076A1 (en) Method for keyword extraction
CN103473327A (zh) 图像检索方法与系统
Suleiman et al. Comparative study of word embeddings models and their usage in Arabic language applications
CN104809142A (zh) 商标查询系统和方法
CN111694946A (zh) 文本关键词可视化显示方法、装置及计算机设备
CN108038099B (zh) 基于词聚类的低频关键词识别方法
Noaman et al. Naive Bayes classifier based Arabic document categorization
CN108363694B (zh) 关键词提取方法及装置
CN112559684A (zh) 一种关键词提取及信息检索方法
Cerra et al. Authorship analysis based on data compression
CN110866102A (zh) 检索处理方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17886995

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17886995

Country of ref document: EP

Kind code of ref document: A1