Similarity score based trademark search method and apparatus

Abstract

A similarity score-based trademark search method performed by an electronic device is disclosed. The similarity score-based trademark search method comprises the steps of: when a target trademark is input, generating a search term corresponding to the target trademark using a first artificial neural network module; inputting the target trademark and the search term into a trademark database to obtain a candidate trademark associated with at least one of the target trademark and the search term; calculating a similarity score between the target trademark and the candidate trademark using a second artificial neural network module; and generating a similarity determination result for the target trademark based on the similarity score. The first artificial neural network module is a module trained to classify the target trademark into components of semantic units and to generate a search term corresponding to the components, and the second artificial neural network module is a module trained to calculate a feature score between the target trademark and the candidate trademark and to calculate a similarity score between the target trademark and the candidate trademark by applying weights to the feature score.

Description

Technology Field

[0001] The present disclosure relates to a trademark search method and apparatus based on a similarity score. Background Technology

[0002] Trademark search technology has been utilized to prevent potential confusion when filing or using a new trademark by examining its similarity to existing registered trademarks. Previously, this was done by simply comparing strings of characters or determining whether there was an exact spelling match.

[0003] However, it has been pointed out that existing methods have limitations in that it is difficult to detect similar trademarks through simple spelling- or pronunciation-based searches, even if they have the same or similar meaning, when the trademarks are expressed through linguistic variations or translations. Additionally, there was a problem of search omissions because the methods failed to detect variations that occur during the actual application process, such as typographical errors or changes in the spelling of loanwords.

[0004] Recently, technologies have been developed to expand the search terms entered into trademark databases to detect a wider variety of candidate trademarks than simple string comparisons, and to determine trademark similarity by utilizing the detected candidate trademarks. Nevertheless, it has been pointed out that technologies based on expanded search suffer from problems such as the excessive extraction of unnecessary candidate trademarks due to an overly broad search scope, or conversely, the failure to identify trademarks with a high probability of actual confusion due to restrictive search term generation rules. The problem to be solved

[0005] One embodiment aims to provide a trademark search method and apparatus that can more efficiently search for various candidate trademarks through decomposition by semantic unit and generation of expanded search terms, without relying on simple spelling comparison when a target trademark is input.

[0006] One embodiment aims to provide a trademark search method and apparatus based on a similarity calculation method capable of precisely evaluating the actual possibility of confusion between trademarks by simultaneously considering pronunciation information and spelling information between a candidate trademark and a target trademark.

[0007] One embodiment aims to provide a trademark search method and apparatus that solves the problem of excessive extraction of unnecessary candidate trademarks during an extended search process and can preferentially identify candidate trademarks with a high probability of confusion.

[0008] One embodiment aims to provide a trademark search method and apparatus that enables a user to quickly and accurately interpret search results by presenting the relationship between trademarks based on similarity scores as an intuitive judgment result, such as risk, caution, and safety.

[0009] However, the problems that the present invention aims to solve are not limited to those mentioned above, and may include problems that are not mentioned but can be clearly understood by those skilled in the art from the description below. means of solving the problem

[0010] A similarity score-based trademark search method performed by an electronic device according to one or more embodiments of the present disclosure comprises: a step of generating a search term corresponding to the target trademark using a first artificial neural network module when a target trademark is input; a step of inputting the target trademark and the search term into a trademark database to obtain a candidate trademark associated with at least one of the target trademark and the search term; a step of calculating a similarity score between the target trademark and the candidate trademark using a second artificial neural network module; and a step of generating a similarity determination result for the target trademark based on the similarity score. The first artificial neural network module is a module trained to classify the target trademark into semantic unit components and to generate a search term corresponding to the components, and the second artificial neural network module is a module trained to calculate a feature score between the target trademark and the candidate trademark and to calculate a similarity score between the target trademark and the candidate trademark by applying weights to the feature score.

[0011] An electronic device according to one or more embodiments of the present disclosure comprises a processor and a memory connected to the processor, wherein the memory is configured to store a program and the processor is configured to execute the program, and when the program is executed, the trademark search method described above is implemented. Effects of the invention

[0012] One embodiment can perform a more comprehensive and precise candidate trademark search than existing technology by automatically generating expanded search terms, such as foreign pronunciations, direct translations, and semantic synonyms, for a target trademark.

[0013] One embodiment can more accurately assess the actual possibility of confusion by calculating a weighted combination of pronunciation feature scores and spelling feature scores between a candidate trademark and a target trademark.

[0014] One embodiment provides the results of the similarity score calculation as intuitive judgment results such as risk, caution, and safety, thereby enabling users to easily understand and utilize trademark similarity judgments.

[0015] One embodiment can improve the efficiency and reliability of search results by automatically identifying the priority among candidate trademarks and selecting the candidate trademark with the highest similarity to generate a final judgment. Brief explanation of the drawing

[0016] FIG. 1 is a drawing for explaining a trademark search process according to one or more embodiments. FIG. 2 is a flowchart illustrating a similarity score-based trademark search method according to one or more embodiments. FIG. 3 is a drawing for explaining a first artificial neural network module according to one or more embodiments. FIG. 4 is a diagram illustrating the process of performing a search of an electronic device according to one or more embodiments. FIG. 5 is a drawing for explaining a second artificial neural network module according to one or more embodiments. FIG. 6 is a drawing for explaining a method for generating a similarity determination result of an electronic device according to one or more embodiments. FIGS. 7 and FIGS. 8 are drawings for explaining a third candidate trademark identification process according to one or more embodiments. FIG. 9 is a diagram illustrating the process of applying a third weight to an electronic device according to one or more embodiments. FIGS. 10 and FIGS. 11 are drawings for explaining a fourth candidate trademark identification process according to one or more embodiments. FIG. 12 is a diagram illustrating the process of generating a similarity determination result of an electronic device according to one or more embodiments. FIG. 13 is a drawing for explaining a trademark search process according to one or more embodiments. FIG. 14 is a block diagram illustrating the configuration of an electronic device according to one or more embodiments. Specific details for implementing the invention

[0017] The various embodiments described in this specification are illustrative for the purpose of clearly explaining the technical concept of this disclosure and are not intended to limit it to specific embodiments. The technical concept of this disclosure includes various modifications, equivalents, alternatives, and embodiments optionally combined from all or part of each embodiment described in this specification. Furthermore, the scope of the technical concept of this disclosure is not limited to the various embodiments presented below or the specific descriptions thereof.

[0018] Terms used in this specification, including technical or scientific terms, may have the meaning generally understood by those skilled in the art to which this disclosure pertains, unless otherwise defined.

[0019] Expressions used herein such as “comprising,” “may compose,” “possessing,” “possessing,” “having,” and “possessing” imply the existence of the subject feature (e.g., function, operation, or component, etc.) and do not exclude the existence of other additional features. That is, such expressions should be understood as open-ended terms implying the possibility of including a second embodiment.

[0020] In this specification, singular expressions include plural expressions unless the context clearly specifies them as singular. Additionally, plural expressions include singular expressions unless the context clearly specifies them as plural. Throughout the specification, when a part is described as including a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0021] Additionally, the terms 'module' or 'part' as used in the specification refer to software or hardware components, and the 'module' or 'part' performs certain roles. However, the meaning of 'module' or 'part' is not limited to software or hardware. The 'module' or 'part' may be configured to reside in an addressable storage medium or configured to run on one or more processors. Thus, as an example, the 'module' or 'part' may include components such as software components, object-oriented software components, class components, and task components, and at least one of processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, or variables. The components and the functions provided within the 'module' or 'part' may be combined into a smaller number of components and 'modules' or 'parts', or further separated into additional components and 'modules' or 'parts'.

[0022] According to one embodiment of the present disclosure, a ‘module’ or ‘part’ may be implemented as a processor and memory. The term ‘processor’ should be broadly interpreted to include a general-purpose processor, a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), a controller, a microcontroller, a state machine, etc. In some environments, the term ‘processor’ may refer to an application-specific integrated circuit (ASIC), a programmable logic device (PLD), a field programmable gate array (FPGA), etc. The term ‘processor’ may also refer to a combination of processing devices, such as, for example, a combination of a DSP and a microprocessor, a combination of multiple microprocessors, a combination of one or more microprocessors combined with a DSP core, or any other combination of such configurations. Additionally, the term ‘memory’ should be broadly interpreted to include any electronic component capable of storing electronic information. 'Memory' may refer to various types of processor-readable media, such as Random Access Memory (RAM), Read-Only Memory (ROM), Non-Volatile Random Access Memory (NVRAM), Programmable Read-Only Memory (PROM), Erasable-Programmable Read-Only Memory (EPROM), Electrically Erasable PROM (EEPROM), Flash Memory, Magnetic or Optical Data Storage Devices, Registers, etc. If a processor can read information from memory and / or write information to memory, the memory is said to be in an electronic communication state with the processor. Memory integrated into a processor is in an electronic communication state with the processor.

[0023] Expressions such as "first," "second," or "first," "second" as used in this specification are used to distinguish one object from another when referring to a plurality of objects of the same kind, unless otherwise indicated in the context, and do not limit the order or importance of said objects.

[0024] Expressions used herein such as “A, B, and C,” “A, B, or C,” “A, B, and / or C,” or “at least one of A, B, and C,” “at least one of A, B, or C,” “at least one of A, B, and / or C,” “at least one selected from A, B, and C,” “at least one selected from A, B, or C,” “at least one selected from A, B, and / or C,” etc., may mean each of the listed items or all possible combinations of the listed items. For example, “at least one selected from A and B” may refer to (1) A, (2) at least one of A, (3) B, (4) at least one of B, (5) at least one of A and at least one of B, (6) at least one of A and B, (7) at least one of B and A, and (8) all of A and B.

[0025] As used herein, the expression “based on” is used to describe one or more factors affecting an act or action of a decision or judgment described in the phrase or sentence containing such expression, and such expression does not exclude additional factors affecting said act or action of a decision or judgment.

[0026] As used in this specification, the expression that a certain component (e.g., a first component) is "connected" or "connected" to another component (e.g., a second component) may mean that the said certain component is not only directly connected or connected to the said other component, but is also connected or connected through a new other component (e.g., a third component).

[0027] As used herein, the expression "configured to" may have meanings such as "set to," "capable of," "modified to," "made to," or "capable of." Such expression is not limited to the meaning of "specifically designed in hardware," and, for example, a processor configured to perform a specific operation may mean a generic-purpose processor capable of performing that specific operation by executing software.

[0028] Various embodiments of the present disclosure will be described below with reference to the accompanying drawings. In the accompanying drawings and the description thereof, identical or substantially equivalent components may be given the same reference numerals. Furthermore, in the description of the various embodiments below, the description of identical or corresponding components may be omitted, but this does not mean that such components are not included in the embodiments.

[0029] FIG. 1 is a drawing for explaining a trademark search process according to one or more embodiments.

[0030] Referring to FIG. 1, according to one embodiment, when a target trademark (10) is input, the electronic device can compare the target trademark (10) with a plurality of trademarks stored in a trademark database to generate a similarity determination result for the target trademark (10). Here, the electronic device can be implemented as various types of electronic devices such as a mobile phone, a smartphone, a personal computer (PC), a tablet PC, a wearable device, an artificial intelligence (AI) device, etc.

[0031] The target trademark (10) may be a search trademark entered by a user to determine trademark similarity based on a trademark database. Before filing or using a trademark, the user may enter the target trademark (10) into an electronic device to compare similarity between multiple trademarks stored in a trademark database. However, the target trademark (10) is not limited thereto and may be referred to in various ways, such as a search trademark, a comparison trademark, or a standard trademark, but in this disclosure, it will be collectively referred to as the target trademark.

[0032] A trademark database stores information on multiple trademarks that have been filed or registered, and may be a set of data used to determine whether there is similarity with a target trademark (10). The trademark database may include various data information such as trademark names, trademark images, and designated goods.

[0033] According to one embodiment, the electronic device may generate a plurality of search terms (20) corresponding to a target trademark (10) using a first artificial neural network module. The first artificial neural network module may be a module trained to classify the target trademark (10) into components of semantic units and to generate search terms corresponding to the components.

[0034] For example, the electronic device can use a first artificial neural network module to generate a search term corresponding to “A” and a search term corresponding to “B” from a target trademark (10) of the form “A+B”, respectively.

[0035] According to one embodiment, the electronic device may input a target trademark (10) and a plurality of generated search terms (20) into a trademark database to obtain candidate trademarks for the target trademark (10). The electronic device may generate a similarity determination result between the target trademark (10) and the candidate trademarks. The electronic device may use a second artificial neural network module to calculate a similarity score between the target trademark (10) and the plurality of candidate trademarks stored in the trademark database. The electronic device may generate a similarity determination result for the target trademark (10) based on the similarity score.

[0036] The similarity determination result may be a comparison result calculated by analyzing whether there is similarity between the target trademark (10) and multiple candidate trademarks stored in the trademark database. The similarity determination result may include a risk determination, a caution determination, and a safety determination. The similarity determination result may be a comparison result calculated based on numerical values ​​for trademark similarity.

[0037] The second artificial neural module may be a module trained to calculate feature scores between a target trademark (10) and a plurality of candidate trademarks stored in a trademark database, and to calculate similarity scores between the target trademark (10) and the plurality of candidate trademarks based on the feature scores.

[0038] For example, the electronic device can use a second artificial neural network module to calculate a similarity score of 8.5 points / 10 points between a target trademark (10) and any one candidate trademark stored in a trademark database, and generate a similarity determination result based on the similarity score.

[0039] The user can input the target trademark (10) to be searched into the trademark database to determine the degree of similarity between the target trademark (10) and multiple trademarks stored in the trademark database. However, since the process of determining similarity based on the trademark database focuses on the degree of similarity in pronunciation or spelling between the target trademark (10) and multiple trademarks, there may be limitations in considering the conceptual and semantic similarity of the trademarks.

[0040] Hereinafter, with reference to the drawings, various embodiments will be described in which an electronic device uses a first artificial neural network module and a second artificial neural network module to generate a similarity determination result between a target trademark (10) and application trademarks.

[0041] FIG. 2 is a flowchart illustrating a similarity score-based trademark search method according to one or more embodiments.

[0042] Referring to FIG. 2, when a target trademark is input, the electronic device can generate a search term corresponding to the target trademark using a first artificial neural network module (S210). The electronic device can generate a search term corresponding to the target trademark using a first artificial neural network module to determine whether a trademark similar or identical to the target trademark is stored in the trademark database.

[0043] The search term may be a string of characters representing spelling variations, extended expressions, or semantically similar expressions of the target trademark. The search term may be an input value entered into a trademark database to search for similar trademarks alike to the target trademark.

[0044] According to one embodiment, the electronic device can acquire candidate trademarks associated with at least one of the target trademark and search term by inputting the target trademark and search term into a trademark database (S220).

[0045] A candidate trademark may be a trademark corresponding to the input result of entering the target trademark and multiple generated search terms into a trademark database. A candidate trademark may be a trademark that includes some components of the target trademark or includes at least one of the multiple search terms. A candidate trademark may be a trademark associated with semantic or visual similarity to at least one of the target trademark and the search terms.

[0046] According to one embodiment, the electronic device can calculate a similarity score between a target trademark and a candidate trademark using a second artificial neural network module (S230).

[0047] The electronic device can calculate feature scores between a target trademark and a candidate trademark using a second artificial neural network module, and calculate a similarity score by applying weights to the feature scores.

[0048] The similarity score may be a numerical value representing the result of comparing features between a target trademark and a candidate trademark. Specifically, the similarity score may be a numerical value representing the degree of similarity regarding appearance, pronunciation, and concept between the target trademark and the candidate trademark. The similarity score is based on a standard of 10 points, and the value of the similarity score may be higher as the target trademark and the candidate trademark become more similar. However, the standard score for the similarity score is merely one example and can be set differently. The similarity score may be referred to by various terms such as feature score, feature distance, or comparison score, but in this disclosure, it will be collectively referred to as the similarity score.

[0049] According to one embodiment, the electronic device can generate a similarity determination result for a target trademark based on a similarity score (S240).

[0050] For example, if the similarity score falls within the risk assessment range, the electronic device determines the similarity assessment result for the target trademark as a risk assessment and can provide a User Interface (UI) containing the assessment result regarding the risk assessment to the user.

[0051] FIG. 3 is a drawing for explaining a first artificial neural network module according to one or more embodiments.

[0052] Referring to FIG. 3, according to one embodiment, an electronic device can use a first artificial neural network module (300) to generate a first extended search term (310) based on the foreign pronunciation of a component, a second extended search term (320) based on a string of direct translations of a component, and the third extended search term (330) based on the semantic similarity of a component.

[0053] For example, an electronic device may receive a target trademark (340) containing components “A,” “B,” and “C.” The electronic device may transmit the target trademark (340) to a first artificial neural network module (300). The first artificial neural network module (300) may classify the target trademark (340) into each of the semantic unit components “A,” “B,” and “C.” The first artificial neural network module (300) may generate and output a first expanded search term (310), a second expanded search term (320), and a third expanded search term (330) corresponding to each component, respectively.

[0054] The first artificial neural network model (300) may be implemented as a Large Language Model (LLM), a pre-trained encoder-based language model, a sequence-to-sequence (seq2seq) conversion model, a multimodal neural network model, etc. However, the embodiment is not necessarily limited thereto, and the artificial neural network model (300) is any machine learning-based model capable of effectively inferring and generating information such as the foreign language pronunciation of a component, a direct translation string, and semantic similarity, and is not limited to a specific model type.

[0055] According to one example, the first extended search term (310) may be a search term based on the foreign pronunciation of each of the multiple components included in the target trademark.

[0056] For example, if the component of the first extended search term (310) is a string composed of Korean characters, the first extended search term (310) may be a search term composed of English based on the foreign pronunciation of the corresponding Korean characters. For example, if the component is “red,” the first extended search term (310) may be “bulgeun” written according to the foreign pronunciation of “red.”

[0057] For example, if the component of the first extended search term (310) is a string composed of English, the first extended search term (310) may be a search term composed of Korean based on the Korean pronunciation of the English. For example, if the component is “red,” the first extended search term (310) may be “Red” written according to the Korean pronunciation of “red”.

[0058] According to one example, the second extended search term (320) may be a search term based on a string of direct translations of each of the multiple components included in the target trademark.

[0059] For example, if the component is a string composed of Korean characters, the second extended search term (320) may be a search term composed of Korean characters or an English word that is a direct translation of the Korean characters. For example, if the component is “red,” the second extended search term (320) may be “red,” which is the English translation of “red.” For example, if the component is “brick,” the English word “brick” corresponding to “brick” is detected, and the second extended search term (320) may be “brick” or “brick,” which is the Korean translation of “brick.”

[0060] For example, if the component is a string composed of English, the second extended search term (320) may be a search term composed of Korean that is a direct translation of the English. For example, if the component is “red,” the second extended search term (320) may be “red” which is a Korean translation of “red”.

[0061] According to one example, the third extended search term (330) may be a search term based on the semantic similarity of each of the multiple components included in the target trademark.

[0062] For example, if the component is a string composed of Korean characters, the third extended search term (330) may be a search term that includes a string semantically similar to the Korean characters. For example, if the component is “red,” the third extended search term (330) may be “red,” “red,” “red,” or “cimson.”

[0063] For example, if the component is a string composed of English, the third extended search term (330) may be a search term containing a string semantically similar to the English. For example, if the component is “brick,” the third extended search term (330) may be “tile” or “building material.”

[0064] FIG. 4 is a diagram illustrating the process of performing a search of an electronic device according to one or more embodiments.

[0065] According to one embodiment, the electronic device can perform searches of different ranges for each of the first expanded search term, the second expanded search term, and the third expanded search term obtained through the first artificial neural network module.

[0066] Referring to FIG. 4, the electronic device can perform a search of the first range by entering the first extended search term into the trademark database (S410).

[0067] The search of the first scope may be a search performed by allowing at least one variation of the spelling and pronunciation of the search term entered into the trademark database. The first expanded search term is a search term based on the pronunciation of a loanword, and depending on the phonetic transcription or the loanword transcription, even if the pronunciation is the same, it may contain different strings or have different spacing positions. That is, the search of the first scope may be a search scope intended to search for all trademarks included in a phonetic similarity relationship beyond literal identity. The search of the first scope may be a search with a broader scope than the search of the second scope.

[0068] For example, if the first extended search term is “red brick,” the electronic device can perform a search of the first range to identify candidate trademarks such as “red brick” and “red beak” that are phonetically similar to “red brick.”

[0069] According to one embodiment, the electronic device can perform a search of the second range by entering at least one of the second extended search term and the third extended search term into the trademark database (S420).

[0070] The search in the second scope is a search scope that searches for candidate trademarks containing the search term itself as is, and may be a narrower scope than the search in the first scope. In the case of the second and third expanded search terms, they are search terms generated by directly translating existing components or based on semantic similarity; if the search scope is performed identically to the search in the first scope, the search scope becomes excessively broad, and trademarks dissimilar to the target trademark may be identified as candidate trademarks. Therefore, the electronic device can identify candidate trademarks by performing a search in the second scope, which is more restrictive than the first scope, on at least one of the second and third expanded search terms.

[0071] According to one embodiment, the electronic device can acquire candidate trademarks based on the search results of a first range and the search results of a second range (S430).

[0072] For example, when a first expanded search term and a second expanded search term are generated through a first artificial neural network module, the electronic device can perform a search of a first range for the first expanded search term and a search of a second range for the second expanded search term based on a trademark database.

[0073] FIG. 5 is a drawing for explaining a second artificial neural network module according to one or more embodiments.

[0074] Referring to FIG. 5, the electronic device can use a second artificial neural network module (500) to calculate feature scores between a target trademark (510) and a candidate trademark (520), and calculate a similarity score (530) based on the feature scores.

[0075] The second artificial neural network module (500) may be a module trained to calculate pronunciation feature scores (540) and spelling feature scores (550) between a target trademark (510) and a candidate trademark (520), and to output a similarity score (530) obtained by averaging the calculated pronunciation feature scores (540) and spelling feature scores (550).

[0076] The pronunciation feature score (540) may be a numerical value obtained by evaluating the phonetic similarity between the target trademark (510) and the candidate trademark (520). Specifically, the pronunciation feature score (540) may be calculated by comprehensively considering the pronunciation flow of the trademark, whether the pronunciation of the preceding part matches, and the similarity when the entire character arrangement is pronounced.

[0077] The spelling feature score (550) may be a numerical value obtained by evaluating the spelling similarity between the target trademark (510) and the candidate trademark (520). The spelling feature score (550) may be calculated based on the number of characters that must be replaced, deleted, or added for the trademark names of the target trademark (510) and the candidate trademark (520) to become identical. A lower value for the spelling feature score (550) indicates that the spelling of the two trademarks is more similar.

[0078] According to one embodiment, the electronic device can obtain a final similarity score between the target trademark (510) and the candidate trademark (520) by applying a pre-set weight to each of the pronunciation feature score (540) and the spelling feature score (550) through the second artificial neural network module (500).

[0079] Meanwhile, the electronic device may perform a preprocessing process on the target trademark (510) and candidate trademark (520) to calculate a more accurate feature score before calculating a similarity score between the target trademark (510) and candidate trademark (520) using a second artificial neural network module.

[0080] According to one embodiment, the electronic device may perform preprocessing on each of the target trademark and the candidate trademark, and calculate a similarity score between the preprocessed target trademark and the preprocessed candidate trademark. Here, the preprocessing may be a process that includes at least one of lowercase conversion processing, space removal processing, and special character removal processing for strings included in the target trademark and the candidate trademark.

[0081] That is, the electronic device can perform preprocessing on each of the two trademarks to calculate feature scores for the target trademark (510) and the candidate trademark (520) under the same or limited conditions, and calculate feature scores by comparing the preprocessed trademarks.

[0082] FIG. 6 is a drawing for explaining a method for generating a similarity determination result of an electronic device according to one or more embodiments.

[0083] According to one embodiment, the electronic device inputs a target trademark and a search term into a trademark database to obtain a plurality of first candidate trademarks based on the original expression of the target trademark and a plurality of second candidate trademarks based on at least one of a first extended search term, a second extended search term, and a third extended search term.

[0084] The first candidate trademark may be a candidate trademark identified based on the original expression of each component included in the subject trademark itself. Specifically, the first candidate trademark may be a candidate trademark identified based on an expression to which minor variations, such as spelling variations or phonetic transcriptions of the subject trademark, have been applied. For example, if the subject trademark is “COFFEJOY,” the first candidate trademark may be “Cafe Joy” or “Coffee Joy” based on the original expression of the subject trademark. That is, the first candidate trademark may be a candidate trademark based on phonetic or spelling variations of the subject trademark.

[0085] The second candidate trademark may be a candidate trademark associated with at least one of the first, second, and third expanded search terms. Specifically, the second candidate trademark may be a candidate trademark generated through the semantic expansion, direct translation, or similar expression of the target trademark. For example, if the target trademark is “REDBRICK,” the second candidate trademark may be “Red Brick,” a direct translation of the target trademark, or “Red Tile,” a semantically expanded version.

[0086] Referring to FIG. 6, the electronic device can calculate feature scores between each of the plurality of first candidate trademarks and the target trademark (S610). The electronic device can calculate feature scores by comparing each of the plurality of first candidate trademarks with the target trademark using a second artificial neural network module.

[0087] According to one embodiment, the electronic device can identify a third candidate trademark with the highest similarity score among a plurality of first candidate trademarks based on feature scores (S620). The electronic device can identify the candidate trademark with the highest feature score, i.e., the highest similarity score among the calculated plurality of feature scores, as the third candidate trademark.

[0088] Specifically, the electronic device can calculate pronunciation feature scores and spelling feature scores among feature scores by comparing each of the plurality of first candidate trademarks with the target trademark, and finally calculate a similarity score by applying weights to each feature score. That is, the electronic device can identify the candidate trademark with the highest final similarity score as the third candidate trademark, rather than identifying the candidate trademark with a high single feature score as the third candidate trademark. An embodiment for calculating a similarity score based on pronunciation feature scores and spelling feature scores will be described in detail in FIG. 7.

[0089] According to one embodiment, the electronic device can calculate feature scores between each of the plurality of second candidate trademarks and the target trademark (S630).

[0090] According to one embodiment, the electronic device can identify a fourth candidate trademark with the highest similarity score among a plurality of second candidate trademarks based on feature scores (S640). Likewise, the electronic device can identify a fourth candidate trademark with the highest similarity score among a plurality of second candidate trademarks in the same manner as the third candidate trademark identification process.

[0091] According to one embodiment, the electronic device can generate a similarity determination result based on the similarity score of the candidate trademark with the higher similarity score among the third candidate trademark and the fourth candidate trademark (S650).

[0092] The electronic device compares the similarity score of the third candidate trademark with the final similarity score of the fourth candidate trademark and can generate a similarity judgment result among a risk judgment, a caution judgment, and a safety judgment based on the similarity score of the candidate trademark with the higher similarity score.

[0093] FIGS. 7 and FIGS. 8 are drawings for explaining a third candidate trademark identification process according to one or more embodiments.

[0094] Referring to FIG. 7, the electronic device can calculate a pronunciation feature score among feature scores based on pronunciation information between each of the plurality of first candidate trademarks and the target trademark (S710). Here, the pronunciation information may be information about the syllables and pronunciation patterns of the target trademark.

[0095] According to one embodiment, the electronic device can calculate a spelling feature score among feature scores based on spelling information between each of the plurality of first candidate trademarks and the target trademark (S720). Here, the spelling information may be information about the character arrangement of the target trademark.

[0096] According to one embodiment, the electronic device can apply a first weight to the pronunciation feature score and a second weight to the spelling feature score to calculate a similarity score for each of the plurality of first candidate trademarks (S730).

[0097] The first weight can be a higher value than the second weight. In determining trademark similarity, phonetic similarity can be more intuitive than spelling similarity, and since people recognize trademarks more often through spoken pronunciation than visual notation, applying a higher weight to phonetic feature scores than to spelling feature scores allows for a more accurate determination of trademark similarity.

[0098] For example, the first weight and the second weight may be values ​​less than 1, and the first weight may be 0.7 and the second weight may be 0.3. However, it goes without saying that the first weight and the second weight may be set differently depending on the device manufacturing stage or user settings.

[0099] According to one embodiment, the electronic device can identify the candidate trademark with the highest similarity score among a plurality of first candidate trademarks as the third candidate trademark (S740).

[0100] Referring to FIG. 8, the electronic device can apply a first weight (830) to the pronunciation feature score (810) for each of the plurality of first candidate trademarks and apply a second weight (840) to the spelling feature score (820) to finally calculate a similarity score (850).

[0101] For example, in the case of trademark A, the electronic device can calculate a similarity score of 7.4 by applying a first weight of 0.7 to a pronunciation feature score of 8.0 and a second weight of 0.3 to a spelling feature score of 6.0.

[0102] For example, in the case of trademark B, the electronic device can calculate a similarity score of 8.47 by applying a first weight of 0.7 to a pronunciation feature score of 9.1 and a second weight of 0.3 to a spelling feature score of 7.0.

[0103] For example, in the case of trademark C, the electronic device can calculate a similarity score of 6.39 by applying a first weight of 0.7 to a pronunciation feature score of 6.0 and a second weight of 0.3 to a spelling feature score of 7.3.

[0104] In this case, the electronic device can identify trademark B, which has the highest similarity score, as the third candidate trademark (860).

[0105] FIG. 9 is a diagram illustrating the process of applying a third weight to an electronic device according to one or more embodiments.

[0106] According to one embodiment, the electronic device can calculate a similarity score for each of the plurality of first candidate trademarks by additionally applying a third weight to the pronunciation feature score of a candidate trademark including a first component among the plurality of first candidate trademarks.

[0107] The first component may be a component of the semantic unit located at the very beginning of the string of characters included in the target trademark. That is, the first component may encompass the overall meaning of the target trademark. For example, if the target trademark is “Red Brick,” the first component may be “Red” because “Red” is located at the beginning of the trademark and defines the overall meaning.

[0108] The third weight may be a factor intended to adjust the importance of the pronunciation feature score in determining trademark similarity. Since the first component is located at the beginning of the target trademark and defines the overall meaning of the trademark, if that component is also included in the candidate trademark, there is a high probability that the two trademarks will be perceived as identical or similar in actual commercial situations. Therefore, by applying the third weight in addition to the pronunciation feature score to determine trademark similarity, the electronic device can generate a more accurate similarity determination result.

[0109] The third weight may be a value of 1 or greater. For example, the third weight may be 1.2, and it goes without saying that it may be set differently depending on the device manufacturing stage or user settings.

[0110] Referring to FIG. 9, if trademark A includes a first component d, the electronic device may additionally apply a third weight of 1.2 to the pronunciation feature score of trademark A. The electronic device may calculate a similarity score of 8.52 by additionally applying the third weight to trademark A. The electronic device may identify trademark A as the third candidate trademark (910) among a plurality of first candidate trademarks (trademark A, trademark B, trademark C).

[0111] FIGS. 10 and FIGS. 11 are drawings for explaining a fourth candidate trademark identification process according to one or more embodiments.

[0112] Referring to FIG. 10, the electronic device can calculate a pronunciation feature score based on pronunciation information between each of the plurality of second candidate trademarks and the target trademark (S1010).

[0113] According to one embodiment, the electronic device can calculate a spelling feature score based on spelling information between each of the plurality of second candidate trademarks and the target trademark (S1020).

[0114] According to one embodiment, the electronic device may apply a fourth weight to the pronunciation feature score and a second weight to the spelling feature score to calculate a similarity score for each of the plurality of second candidate trademarks (S1030).

[0115] The fourth weight may be a weight established to reflect the characteristics identified by the second candidate trademark based on expanded search terms (first expanded search term, second expanded search term, third expanded search term), unlike the first candidate trademark. The first candidate trademark is a trademark identified based on the original expression of the target trademark; since there is a high likelihood of confusion when phonetic similarity is high, the first weight, which places significant weight on the phonetic feature score, may be applied accordingly.

[0116] On the other hand, since the second candidate trademark is identified based on expanded search terms generated through translation, semantic expansion, etc., pronunciation similarity may not be as direct as that of the first candidate trademark. Therefore, unlike the third candidate trademark identification process, the electronic device may apply a fourth weight different from the first weight to adjust the influence of the pronunciation feature score in the fourth candidate trademark identification process.

[0117] For example, if the target trademark is “Red Brick” and the second candidate trademark is “Red Brick,” the two trademarks may be closely related in meaning but differ significantly in pronunciation. The electronic device can prevent the second candidate trademark, based on semantic expansion, from being determined to be excessively similar to the target trademark by adjusting the weight of the pronunciation feature score through the application of a fourth weight different from the third weight to the second candidate trademark.

[0118] The fourth weight may be lower than the first weight and higher than the second weight. For example, the fourth weight may be 0.6, and it goes without saying that it may be set differently depending on the device manufacturing stage or user settings.

[0119] According to one embodiment, the electronic device can identify the candidate trademark with the highest similarity score among a plurality of second candidate trademarks as the fourth candidate trademark (S1040).

[0120] Referring to FIG. 11, the electronic device can apply a fourth weight (1130) to the pronunciation feature score (1110) for each of the plurality of second candidate trademarks and apply a second weight (1140) to the spelling feature score (1120) to finally calculate a similarity score (1150).

[0121] For example, in the case of trademark E, the electronic device can calculate a similarity score of 5.4 by applying a fourth weight of 0.6 to a pronunciation feature score of 7.0 and a second weight of 0.3 to a spelling feature score of 4.0.

[0122] For example, in the case of trademark F, the electronic device can calculate a similarity score of 4.05 by applying a fourth weight of 0.6 to a pronunciation feature score of 5.0 and a second weight of 0.3 to a spelling feature score of 3.5.

[0123] For example, in the case of trademark G, the electronic device can calculate a similarity score of 5.25 by applying a fourth weight of 0.6 to a pronunciation feature score of 6.5 and a second weight of 0.3 to a spelling feature score of 4.0.

[0124] In this case, the electronic device can identify trademark E, which has the highest similarity score, as the fourth candidate trademark (1160).

[0125] FIG. 12 is a diagram illustrating the process of generating a similarity determination result of an electronic device according to one or more embodiments.

[0126] Referring to FIG. 12, the electronic device can identify the similarity score of the candidate trademark with the higher similarity score among the third candidate trademark and the fourth candidate trademark (S1210). For example, if the similarity score of the third candidate trademark is 8.52 and the similarity score of the fourth candidate trademark is 5.4, the electronic device can identify the similarity score of the third candidate trademark with the higher similarity score of 8.52.

[0127] According to one embodiment, the electronic device can generate a similarity judgment result corresponding to the similarity score of a candidate trademark with a higher similarity score based on a similarity judgment table for each similarity score interval.

[0128] The similarity judgment table for each similarity score range may be a data table in which similarity judgment results corresponding to pre-set similarity score ranges are mapped. For example, it may be a data table in which similarity scores 0 to 4 are mapped as safe judgment, 4 to 7 as caution judgment, and 7 to 10 as danger judgment. However, this is merely one example, and it goes without saying that the similarity judgment table for each similarity score range may be configured differently depending on the device manufacturing stage or user settings.

[0129] According to one embodiment, if a third candidate trademark is identified as a candidate trademark with a higher similarity score, the electronic device can generate a similarity judgment result based on the higher feature score among the pronunciation feature score and the spelling feature score of the third candidate trademark, based on a similarity judgment table by similarity score interval.

[0130] Referring again to FIG. 9, the electronic device can identify that trademark A, identified as a third candidate trademark, has a pronunciation feature score of 9.6 and a spelling feature score of 6.0. The electronic device can generate a similarity determination result based on the higher of the pronunciation feature score and the spelling feature score for trademark A, which is 9.6. If the similarity determination table is as in the example described above, the electronic device can generate a similarity determination result for the target trademark as a risk determination.

[0131] According to one embodiment, if the electronic device identifies the fourth candidate trademark as the candidate trademark with a higher similarity score, it can generate a similarity judgment result based on the lower feature score between the pronunciation feature score and the spelling feature score of the fourth candidate trademark based on a similarity judgment table by similarity score interval (S1230).

[0132] Referring again to FIG. 11, the electronic device can identify that trademark E, identified as the fourth candidate trademark, has a pronunciation feature score of 7.0 and a spelling feature score of 4.0. The electronic device can generate a similarity determination result based on the pronunciation feature score and spelling feature score of trademark E, with the lower feature score being 4.0. If the similarity determination table is as in the example described above, the electronic device can generate a similarity determination result for the target trademark as a safe determination.

[0133] In the case of a third candidate trademark based on the original expression of the subject trademark, the likelihood of confusion with the subject trademark is relatively high; therefore, the electronic device may perform a similarity assessment between the subject trademark and the candidate trademark by applying a more lenient standard. On the other hand, in the case of a fourth candidate trademark based on expanded search terms, direct similarity with the subject trademark is low and may include semantic extensions or variations; therefore, the electronic device may perform a similarity assessment by applying a stricter baseline to prevent an excessive determination of similarity with the subject trademark.

[0134] According to one embodiment, the electronic device can identify one of the risk judgment, caution judgment and safety judgment among the similarity judgment results based on the similarity score (S1240).

[0135] According to one embodiment, if the identified judgment result is either a risk judgment or a caution judgment, the electronic device may additionally output a replacement trademark included in the similarity range corresponding to the safety judgment. That is, the electronic device may provide the user with the trademark having the lowest similarity score among a plurality of third candidate trademarks and a plurality of fourth candidate trademarks as a replacement trademark.

[0136] FIG. 13 is a drawing for explaining a trademark search process according to one or more embodiments.

[0137] Referring to FIG. 13, when a target trademark is entered, the electronic device can identify a plurality of candidate trademarks (1310) corresponding to the target trademark and search term based on a trademark database.

[0138] The electronic device can identify a third candidate trademark and a fourth candidate trademark among a plurality of candidate trademarks (1310) and identify a similarity score between the target trademarks. The electronic device can generate a similarity determination result based on the identified similarity score and provide a UI corresponding to the similarity determination result to the user.

[0139] Since the method for identifying similarity scores and generating similarity judgment results has been specifically explained through the embodiments described above, a detailed explanation thereof will be omitted.

[0140] FIG. 14 is a block diagram illustrating the configuration of an electronic device according to one or more embodiments.

[0141] Referring to FIG. 14, the electronic device (100) includes a memory (1410) and a processor (1420). However, it is not limited thereto, and the electronic device (100) may be implemented with some components excluded or with other components included.

[0142] The memory (1410) can store at least one instruction, data, program, etc. required for the operation of the electronic device (100). Depending on the purpose of data storage, the memory (1410) may be implemented in the form of a memory embedded in the electronic device (100) or in the form of a memory that is detachable from the electronic device (100). For example, data for operating the electronic device (100) may be stored in a memory embedded in the electronic device (100), and data for the expansion function of the electronic device (100) may be stored in a memory that is detachable from the electronic device (100).

[0143] The processor (1420) controls the overall operation of the electronic device (100). Specifically, the processor (1420) is connected to each component of the electronic device (100) to control the overall operation of the electronic device (100). The processor (1420) may be configured to execute a program stored in memory (1410).

[0144] According to one embodiment, the electronic device (100) can implement a method of calculating a similarity score and generating a similarity determination result based on the calculated similarity score when a program stored in memory (1410) is executed.

[0145] It is obvious that each step or operation of the method according to the embodiments of the present disclosure may be performed by a computer comprising one or more processors in accordance with the execution of a computer program stored in a computer-readable recording medium.

[0146] The computer-executable instructions stored on the aforementioned recording medium can be implemented through a computer program programmed to perform each corresponding step, and such a computer program can be stored on a computer-readable recording medium and executed by a processor. The computer-readable recording medium may be a non-transitory readable medium. In this case, a non-transitory readable medium refers to a medium that stores data semi-permanently and is readable by a device, rather than a medium that stores data for a short moment, such as a register, cache, or memory. Specifically, programs for performing the various methods described above may be provided by being stored on a non-transitory readable medium, such as semiconductor memory devices including erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), and flash memory devices; magnetic disks such as internal hard disks and removable disks; optical-magnetic disks; and non-volatile memory including CD-ROMs and DVD-ROMs.

[0147] Methods according to the various examples disclosed in this document may be provided by being included in a computer program product. The computer program product may be distributed in the form of a device-readable storage medium (e.g., compact disc read-only memory (CD-ROM)) or online through an application store (e.g., Play Store™). In the case of online distribution, at least a portion of the computer program product may be temporarily stored or temporarily created on a storage medium, such as the memory of a manufacturer's server, an application store's server, or a relay server.

[0148] As explained above, a person skilled in the art to which this disclosure pertains will understand that this disclosure may be implemented in other specific forms without altering its technical concept or essential features. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. The scope of this disclosure is defined by the claims set forth below rather than by the detailed description, and all modifications or variations derived from the meaning and scope of the claims and equivalent concepts should be interpreted as being included within the scope of this disclosure.

[0149] The features and advantages described herein are not all included, and in particular, many additional features and advantages will become apparent to those skilled in the art by considering the drawings, the specification, and the claims. Furthermore, it should be noted that the language used in this specification has been chosen primarily for readability and instructional purposes and may not be chosen to describe or limit the subject matter of this disclosure.

[0150] The foregoing description of the embodiments of the present disclosure is provided for illustrative purposes only. It is not intended to limit the present disclosure to the exact form disclosed or to make it incomplete. Those skilled in the art will understand that many modifications and variations are possible in light of the foregoing disclosure.

[0151] Therefore, the scope of the present disclosure is not limited by the detailed description but by any of the claims of the application based thereon. Accordingly, the disclosure of embodiments of the present disclosure is illustrative and does not limit the scope of the present disclosure as set forth in the following claims.

Claims

Claim 1 A similarity score-based trademark search method performed by an electronic device, comprising: a step of, when a target trademark is input, generating a search term using a first artificial neural network module including a first extended search term based on the foreign pronunciation of a component of the target trademark corresponding to the target trademark, a second extended search term based on a string of direct translation of the component, and a third extended search term based on the semantic similarity of the component; a step of inputting the target trademark and the search term into a trademark database to obtain a plurality of first candidate trademarks based on the original expression of the target trademark and a plurality of second candidate trademarks based on at least one of the first extended search term, the second extended search term, and the third extended search term; a step of using a second artificial neural network module to calculate a similarity score between each of the plurality of first candidate trademarks and the target trademark to identify the third candidate trademark with the highest similarity score; and a step of using the second artificial neural network module to calculate a similarity score between each of the plurality of second candidate trademarks and the target trademark to identify the fourth candidate trademark with the highest similarity score. A trademark search method comprising: a step of generating a similarity determination result for the target trademark based on the similarity score of the candidate trademark with the higher similarity score among the third candidate trademark and the fourth candidate trademark; wherein the first artificial neural network module is a module trained to classify the target trademark into components of semantic units and generate search terms corresponding to the components, and the second artificial neural network module is a module trained to calculate a feature score between the target trademark and the candidate trademark and to calculate a similarity score between the target trademark and the candidate trademark by applying weights to the feature score. Claim 2 delete Claim 3 A trademark search method according to claim 1, wherein the step of acquiring the candidate trademark comprises: a step of performing a first-range search by inputting the first expanded search term into the trademark database; a step of performing a second-range search by inputting at least one of the second expanded search term and the third expanded search term into the trademark database; and a step of acquiring the candidate trademark based on the search results of the first range and the search results of the second range; wherein the search of the first range is a search performed by allowing at least one variation of the spelling and pronunciation of the search term input into the trademark database, and is a search of a broader range than the search of the second range. Claim 4 A trademark search method according to claim 1, wherein the step of calculating a similarity score between the target trademark and the candidate trademark using the second artificial neural network module comprises: a step of performing preprocessing on each of the target trademark and the candidate trademark; and a step of calculating a similarity score between the preprocessed target trademark and the preprocessed candidate trademark; wherein the preprocessing comprises at least one of lowercase conversion processing, space removal processing, and special character removal processing for strings included in the target trademark and the candidate trademark. Claim 5 A trademark search method according to claim 1, wherein the step of generating a similarity determination result for the target trademark based on the similarity score of the candidate trademark with the higher similarity score among the third candidate trademark and the fourth candidate trademark comprises: a step of calculating a feature score between each of the plurality of first candidate trademarks and the target trademark, and identifying the third candidate trademark with the highest similarity score among the plurality of first candidate trademarks based on the feature score; and a step of calculating a feature score between each of the plurality of second candidate trademarks and the target trademark, and identifying the fourth candidate trademark with the highest similarity score among the plurality of second candidate trademarks based on the feature score. Claim 6 A trademark search method according to claim 5, wherein the step of identifying the third candidate trademark comprises: a step of calculating a pronunciation feature score among the feature scores based on pronunciation information between each of the plurality of first candidate trademarks and the target trademark; a step of calculating a spelling feature score among the feature scores based on spelling information between each of the plurality of first candidate trademarks and the target trademark; a step of applying a first weight to the pronunciation feature score and applying a second weight to the spelling feature score to calculate a similarity score for each of the plurality of first candidate trademarks; and a step of identifying the candidate trademark with the highest similarity score among the plurality of first candidate trademarks as the third candidate trademark; wherein the first weight is a higher value than the second weight. Claim 7 A trademark search method according to claim 6, wherein the target trademark includes a first component of a semantic unit located at the very beginning of the string of trademarks, and the step of calculating a similarity score for each of the plurality of first candidate trademarks by applying a first weight to the pronunciation feature score and applying a second weight to the spelling feature score, further comprising the step of calculating a similarity score for each of the plurality of first candidate trademarks by additionally applying a third weight to the pronunciation feature score of a candidate trademark including the first component among the plurality of first candidate trademarks. Claim 8 A trademark search method according to claim 7, wherein the step of identifying the fourth candidate trademark comprises: a step of calculating the pronunciation feature score based on the pronunciation information between each of the plurality of second candidate trademarks and the target trademark; a step of calculating the spelling feature score based on the spelling information between each of the plurality of second candidate trademarks and the target trademark; a step of applying a fourth weight to the pronunciation feature score and applying the second weight to the spelling feature score to calculate a similarity score for each of the plurality of second candidate trademarks; and a step of identifying the candidate trademark with the highest similarity score among the plurality of second candidate trademarks as the fourth candidate trademark; wherein the fourth weight is a value lower than the first weight and higher than the second weight. Claim 9 A trademark search method according to claim 6, wherein the pronunciation information is information regarding the syllables and pronunciation patterns of the target trademark, and the spelling information is information regarding the character arrangement of the target trademark. Claim 10 A trademark search method according to claim 8, wherein the step of generating the similarity determination result comprises: a step of identifying the similarity score of the candidate trademark with the higher similarity score among the third candidate trademark and the fourth candidate trademark; and a step of generating a similarity determination result corresponding to the similarity score of the candidate trademark with the higher similarity score based on a similarity determination table by similarity score interval. Claim 11 A trademark search method according to claim 10, wherein the step of generating the similarity determination result comprises: a step of generating the similarity determination result based on the similarity determination table by similarity score interval, wherein, when the third candidate trademark is identified as the candidate trademark with the higher similarity score, the higher feature score among the pronunciation feature score and the spelling feature score of the third candidate trademark; and a step of generating the similarity determination result based on the similarity determination table by similarity score interval, wherein, when the fourth candidate trademark is identified as the candidate trademark with the higher similarity score, the lower feature score among the pronunciation feature score and the spelling feature score of the fourth candidate trademark. Claim 12 A trademark search method according to claim 1, wherein the similarity determination result includes a risk determination, a caution determination, and a safety determination, and the step of generating the similarity determination result comprises: a step of identifying any one of the risk determination, the caution determination, and the safety determination among the similarity determination results based on the similarity score; a step of outputting the identified determination result; and a step of additionally outputting an alternative trademark included in a similarity interval corresponding to the safety determination when the identified determination result is any one of the risk determination and the caution determination. Claim 13 An electronic device comprising a processor and a memory connected to the processor, wherein the memory is configured to store a program and the processor is configured to execute the program, and when the program is executed, any one of the methods of claims 1 and 3 through 12 is implemented.

Images