KR20220109955A - Ommited - Google Patents

Abstract

Provided is a method for determining a similarity of a trademark. The method for determining a similarity of a trademark includes: a step of acquiring, by a server, a similar trademark by inputting a target trademark to first deep learning; a step of acquiring, by the server, a plurality of words included in the target trademark and acquiring a plurality of words included in the similar trademark; a step of acquiring, by the server, a first word among the plurality of words included in the target trademark and acquiring a second word among the plurality of words included in the similar trademark; and a step of inputting, by the server, the first word and the second word to second deep learning to acquire an evaluation result for the first word and the second word.

Description

Method, apparatus and system for determining similarity of trademarks using deep learning {OMMITED}

The present invention relates to a method, apparatus and system for determining the similarity of a trademark using deep learning.

With the development of the 4th industry, the value of intellectual property rights is increasing. Accordingly, many people are increasingly interested in trademark applications to protect their own brand names as well as personal brands.

Meanwhile, in order to apply for a trademark, a prior brand name search is performed to determine whether one's own brand name can receive a trademark, and a prior brand name search can be performed by searching various previously published trademark literature.

The most important thing in conducting prior brand name research is to determine whether there is a similar trademark that can deny the inventive step of the target trademark.

However, the amount of trademark documents is huge and it is virtually impossible to analyze all trademark documents published in the past due to time constraints. It is a reality to acquire similar trademarks.

However, the acquisition of similar trademarks through search-type input, etc. is often dependent on the ability of personnel to conduct prior brand name research, and thus, in many cases, stable results of prior brand name research are not guaranteed.

Therefore, the need for a prior brand name research method that can guarantee stable results with a small investment of time is emerging.

Republic of Korea Patent Publication No. 10-2018-0110713 (2018.10.11)

The problem to be solved by the present invention is to provide a method, apparatus and system for determining the similarity of a trademark using deep learning.

The problems to be solved by the present invention are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by a typical brand name from the description below.

A method for determining the degree of similarity of a trademark according to an aspect of the present invention for solving the above-described problem includes, by a server, inputting a target trademark into a first deep learning to obtain a similar trademark; acquiring, by the server, a plurality of words included in the target trademark, and acquiring a plurality of words included in the similar trademark; obtaining, by the server, a first word from among a plurality of words included in the target trademark, and acquiring a second word from a plurality of words included in the similar trademark; inputting, by the server, the first word and the second word into second deep learning, obtaining evaluation results for the first word and the second word; includes

In this case, the first deep learning is CNN-based deep learning learned by inputting a plurality of trademarks as learning data, and the second deep learning corresponds to a trademark, a prior research report matched with the trademark document, and the prior research report It may be deep learning based on Bi-LSTM model trained based on similar trademarks.

In this case, the acquiring of the similar trademark may include: acquiring a plurality of word sets included in the target trademark; obtaining a plurality of target brand words by modeling the obtained plurality of word sets; Acquire a plurality of midpoints for each of the plurality of target brand words, and determine the location of the target trademark based on the plurality of midpoints for each of the plurality of target brand words and the number of word sets included in the target brand word to do; obtaining a plurality of word sets included in a trademark, and modeling the obtained plurality of word sets to obtain a plurality of trademark words; obtaining a plurality of midpoints for each of the plurality of trademark words, and determining the position of the trademark based on the plurality of midpoints for each of the plurality of trademark words and the number of word sets included in the trademark words; and when the location of the target trademark and the location of the trademark are within a preset distance, determining the trademark as the similar trademark; may include.

In this case, the acquiring of the plurality of words may include: acquiring a plurality of word sets for each of the plurality of words included in the target trademark and the similar trademark; obtaining a plurality of importance scores for each of the obtained plurality of word sets; determining an importance score of a word included in the target trademark based on the plurality of importance scores; and arranging the plurality of words in order of importance based on the importance score of each of the plurality of words. may include.

In this case, the steps of obtaining a first word from among the plurality of words included in the target trademark and acquiring a second word from among the plurality of words included in the similar trademark include assigning the plurality of words included in the target trademark to an importance score classifying into n preset groups based on classifying the plurality of words included in the similar trademark into n preset groups based on the importance score; when the first word is included in a first group among n groups of the target trademark, acquiring the second word as all words included in the similar trademark; When the first word is included in the kth group among the n groups of the target trademark, the second word is included in the first group to the (n-k+1)th group among the n groups of the similar trademark. Including; obtaining as a word; wherein n may be a natural number greater than or equal to 1, and k may be a natural number less than or equal to n.

In this case, the obtaining of the evaluation results for the first word and the second word may include: obtaining a similarity score and a dissimilarity score of the first word and the second word, respectively; determining that the first word and the second word are unrelated words when the dissimilarity score is equal to or greater than a preset score; acquiring at least one word set not included in the second word among the word sets included in the first word when the similarity score is equal to or greater than a preset score; obtaining at least an importance score for each of at least one word set not included in the second word, and determining whether a word set equal to or greater than a preset importance score exists among the at least one obtained importance score; determining the first word and the second word as a matching word when there is no word set equal to or greater than a preset importance score; and determining that the first word and the second word are mismatched words when there is a set of words equal to or greater than the preset importance score. may include.

In this case, the obtaining of the word set may include: obtaining a plurality of word sets by analyzing morphemes of the target trademark; determining an erroneous word set among the obtained plurality of word sets; correcting the error word set; and obtaining at least one compound noun phrase set based on the degree of association between the plurality of word sets, wherein the determining of the erroneous word set includes semantic information for each of the obtained word sets. matching; obtaining an accuracy score for each of the plurality of word sets based on the matched semantic information and the trademark; and determining a word set whose accuracy score is equal to or less than a preset value as an erroneous word set, wherein the correcting of the erroneous word set includes: obtaining a plurality of semantic information for the erroneous word; obtaining a plurality of weights for each of the obtained pieces of semantic information; obtaining a weight having the highest degree of relevance with the trademark among the plurality of weights; and matching semantic information corresponding to a weight having the highest degree of relevance with the trademark to the erroneous word set. including, wherein the obtaining of the at least one compound noun phrase set includes: obtaining a plurality of word sets for words included in the trademark; obtaining a plurality of compound noun phrase set candidates obtained by combining the plurality of word sets; obtaining a frequency in which a complex noun phrase set identical to the obtained plurality of compound noun phrase set candidates is included in a trademark; and determining a compound noun phrase set candidate whose frequency is equal to or greater than a preset frequency as a compound noun phrase set. and, the compound noun phrase set candidate may include order information and spacing information of word sets included in the compound noun phrase set candidate.

In this case, the step of obtaining the importance score may include: obtaining, by the server, a word to be calculated for the importance score from the target trademark; The server determines, by the server, the first detailed importance of the word in the entire trademark, the second detailed importance of the word in the trademark classification information corresponding to the similar information of the target trademark, and the search trademark including the word among all the trademarks. calculating one or more detailed importance levels from among the third detailed importance levels; and obtaining, by the server, the importance score of the word based on one or more of the first detailed importance level, the second detailed importance level, and the third detailed importance level; may include.

Other specific details of the invention are included in the detailed description and drawings.

According to the above-described embodiment of the present invention, a prior trademark research method that can be guaranteed a stable result with a small time investment using deep learning can be provided.

Effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the following description.

1 is an exemplary diagram for explaining a system according to an embodiment of the present invention.
2 is a flowchart illustrating a method of obtaining evaluation results for words included in a target trademark and a similar trademark according to an embodiment of the present invention.
3 is a flowchart illustrating a method of obtaining a similar trademark according to an embodiment of the present invention.
4 is a flowchart for explaining a method of obtaining a plurality of words from a target trademark and a similar trademark and arranging them in order of importance according to an embodiment of the present invention.

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only these embodiments allow the disclosure of the present invention to be complete, and the general purpose of the brand name field to which the present invention belongs It is provided to fully inform the brand name of the scope of the present invention, and the present invention is only defined by the scope of the claims.

Unless otherwise defined, all terms (including brand names and scientific terms) used herein may be used with meanings commonly understood by those of ordinary skill in the brand name field to which the present invention belongs. In addition, terms defined in a commonly used dictionary are not to be interpreted ideally or excessively unless clearly specifically defined.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

Each step described in this specification is described as being performed by a computer, but the subject of each step is not limited thereto, and at least a portion of each step may be performed in different devices according to embodiments.

1 is an exemplary diagram for explaining a system according to an embodiment of the present invention.

The system for determining the similarity of a trademark according to the present invention includes a server 10 and an electronic device 20 .

The server 10 is configured to acquire a target trademark, obtain a similar trademark from the acquired target trademark, and determine a degree of similarity between a word of the target trademark and a word of a similar trademark.

Specifically, the server 10 may receive a target trademark from the electronic device 20 or acquire a target trademark from an external server.

In the present specification, a trademark is a concept including a target trademark and a similar trademark, and may be a document about the content of a brand name submitted by an applicant to obtain trademark registration in each country's trademark office. However, the present invention is not limited thereto, and the trademark may be understood as a concept including various documents including the contents of the brand name, such as a job invention for a trademark application, and a thesis. According to an embodiment, the target trademark may be at least one of a job invention book and a thesis for a trademark application, and the similar trademark may be at least one of a job invention book, a thesis, and a trademark application for a trademark application.

The electronic device 20 is configured to provide a trademark to the server 10 . The electronic device 200 according to the present invention may be implemented as a smart phone, but this is only an embodiment, and a smart phone, a tablet personal computer (PC), a mobile phone, a video phone, e-book reader, desktop PC, laptop PC, netbook computer, workstation, server, PDA (personal digital assistant), PMP (portable multimedia) player) or a wearable device.

2 is a flowchart illustrating a method of obtaining evaluation results for words included in a target trademark and a similar trademark according to an embodiment of the present invention.

In step S110, the server 10 may obtain a similar trademark by inputting the target trademark into the first deep learning.

Specifically, the server 10 may obtain a similar trademark similar to the target trademark from among a plurality of trademarks stored in the first database by using the first deep learning.

In one embodiment, the server 10 analyzes each of the plurality of trademarks based on the morpheme analysis of each of the plurality of trademarks stored in the first database, and analyzes the target trademark based on the morpheme analysis of the target trademark. After that, it is possible to obtain a similar trademark with a high degree of relevance to the target trademark. A specific method of obtaining a similar trademark will be described later with reference to FIG. 3 .

In step S120 , the server 10 may acquire a plurality of words included in the target trademark and acquire a plurality of words included in the similar trademark.

In step S130 , the server 10 may acquire a first word among a plurality of words included in the target trademark, and acquire a second word among a plurality of words included in the similar trademark.

In step S140 , the server 10 may input the first word and the second word into a second deep learning to obtain evaluation results for the first word and the second word.

On the other hand, the first deep learning is a deep learning based on a convolutional deep neural network (CNN) that is learned by inputting a plurality of trademarks as learning data, and the second deep learning is a trademark, prior research matching the trademark document It may be a deep learning based on a Bi-LSTM model learned based on a report and a similar trademark corresponding to the prior research report. However, the present invention is not limited thereto, and it goes without saying that various deep learning methods may be applied to the present invention. For example, a model such as a Deep Neural Network (DNN), a Recurrent Neural Network (RNN), or a Bidirectional Recurrent Deep Neural Network (BRDNN) may be used as the deep learning, but is not limited thereto.

In this case, convolutional deep neural networks (CNNs) are a type of multilayer perceptrons designed to use minimal preprocessing. A convolutional neural network consists of one or several convolutional layers and general artificial neural network layers on top of it, and additionally utilizes weights and pooling layers. Thanks to this structure, the convolutional neural network can fully utilize the input data of the two-dimensional structure. In addition, convolutional neural networks can be trained via standard backpropagation. Convolutional neural networks are easier to train than other feed-forward neural network techniques and have the advantage of using fewer parameters. A convolutional neural network extracts features from an input image by alternately performing convolution and subsampling on the input image.

A convolutional neural network includes several convolution layers, several subsampling layers (Subsampling layer, Lacal pooling layer, Max-Pooling layer), and a fully connected layer. The convolution layer is a layer that performs convolution on an input image. In addition, the subsampling layer is a layer that locally extracts the maximum value from the input image and maps it to a 2D image, and subsampling is performed by enlarging the local area.

In the convolution layer, information such as the kernel size, the number of kernels to be used (ie, the number of maps to be created), and a weight table to be applied during the convolution operation are required. For example, suppose that the size of the input image is 32x32, the size of the kernel is 5x5, and the number of kernels to be used is 20. In this case, if a 5×5 kernel is applied to a 32×32 input image, it is impossible to apply the kernel to two pixels above, below, left, and right of the input image. When the kernel is placed on the input image and then convolution is performed, the result value of '-8' is determined as the value of the pixel corresponding to the center element of the kernel among the pixels of the input image included in the kernel. to be. Therefore, when convolution is performed by applying a 5×5 kernel to a 32×32 input image, a 28×28 map is generated. Previously, since it was assumed that the total number of kernels to be used is 20, in the first convolutional layer, a total of 20 maps with a size of 28×28 are generated.

In the subsampling layer, information on the size of a kernel to be subsampled and information on whether to select a maximum value or a minimum value among values in the kernel region are required.

In addition, deep neural networks (DNNs) are artificial neural networks (ANNs) composed of a plurality of hidden layers between an input layer and an output layer.

In this case, the structure of the deep neural network may be composed of a perceptron. A perceptron consists of several inputs, one processor, and one output value. The processor multiplies each input value by a weight, and then sums the input values multiplied by the weight. Then, the processor outputs one output value by substituting the summed value into the activation function. If a specific value is desired as the output value of the activation function, the weight multiplied by each input value may be corrected, and the output value may be recalculated using the modified weight. In this case, each perceptron may use a different activation function. In addition, each perceptron receives the outputs from the previous layer as input, and then uses the activation function to obtain the output. The obtained output is transferred to the input of the next layer. Through the process as described above, some output values can be finally obtained.

In addition, returning to the description of the deep learning technique, a Reccurent Neural Network (RNN) refers to a neural network in which connections between units constituting an artificial neural network constitute a directed cycle. Unlike forward neural networks, recurrent neural networks can utilize the memory inside the neural network to process arbitrary inputs.

Deep Belief Networks (DBN) is a generative graphical model used in machine learning, and in deep learning, it means a deep neural network composed of multiple layers of latent variables. There is a connection between layers, but there is no connection between units within a layer.

The deep trust neural network can be used for prior learning due to the nature of the generative model, and after learning the initial weights through prior learning, the weights can be fine-tuned through backpropagation or other discrimination algorithms. This characteristic is very useful when the training data is small, because the smaller the training data, the stronger the influence of the initial value of the weight on the resulting model. The pre-learned initial weight value is closer to the optimal weight compared to the arbitrarily set initial weight value, which enables performance and speed improvement in the fine-tuning stage.

The above-described artificial intelligence and its learning method have been described for illustrative purposes, and the artificial intelligence and its learning method used in the embodiments described below are not limited. For example, all kinds of artificial intelligence brand names that can be applied to common brand names in the art to solve the same problem and their learning methods can be utilized to implement the system according to the disclosed embodiment.

3 is a flowchart illustrating a method of obtaining a similar trademark according to an embodiment of the present invention.

In step S210, the server 10 may obtain a plurality of word sets included in the target trademark.

In step S220, the server 10 may obtain a plurality of target brand words by modeling the obtained plurality of word sets.

In this case, various modeling techniques may be applied to modeling according to the present invention. In an embodiment, the server 10 may acquire a plurality of target brand words by using an Elasticsearch clustering modeling technique. In another embodiment, the server 10 uses at least one modeling technique among K-means modeling, DBSCAN (Density-based spatial clustering of applications with noise) modeling, hierarchical modeling, and mixed Gaussian modeling techniques to obtain a plurality of targets Trademark words can be obtained.

In step S230, the server 10 obtains a plurality of midpoints for each of the plurality of target brand words, a plurality of midpoints for each of the plurality of target brand words, and the number of word sets included in the target brand words It is possible to determine the position of the target trademark based on the .

In one embodiment, the midpoint of the target brand word may be obtained based on the positions of a plurality of word sets included in the corresponding word, and the target brand position may be the center of gravity with respect to the midpoints of the plurality of target brand words have.

In step S240 , the server 10 may acquire a plurality of word sets included in a trademark, and obtain a plurality of trademark words by modeling the acquired plurality of word sets.

In step S250, the server 10 obtains a plurality of midpoints for each of the plurality of brand words, based on the plurality of midpoints for each of the plurality of brand words and the number of word sets included in the brand words It is possible to determine the position of the trademark.

In an embodiment, the midpoint of the trademark word may be obtained based on positions of a plurality of word sets included in the corresponding word, and the location of the trademark may be a center of gravity with respect to the midpoints of the plurality of trademark words.

In step S260, when the location of the target trademark and the location of the trademark are within a preset distance, the server 10 may determine the trademark as the similar trademark.

In one embodiment, the server 10 may determine a plurality of trademarks existing within a preset distance from the location of the target trademark as similar trademarks, and sort the plurality of similar trademarks determined in the order of similar trademarks having a close distance to the target trademark. can do.

4 is a flowchart for explaining a method of obtaining a plurality of words from a target trademark and a similar trademark and arranging them in order of importance according to an embodiment of the present invention.

In step S310, the server 10 may obtain a plurality of word sets for each of a plurality of words included in the target trademark and the similar trademark.

In step S320 , the server 10 may obtain a plurality of importance scores for each of the obtained plurality of word sets. A method of obtaining the importance score will be described later.

In step S330, the server 10 may determine the importance score of the word included in the target trademark based on the plurality of importance scores.

In an embodiment, the server 10 may obtain a degree of association between a plurality of word sets and a target trademark, and determine the sum of the importance scores of the word sets having the obtained association degree equal to or greater than a preset value as the importance score of the word. have. In this case, the degree of association may be obtained through various methods. For example, the degree of association between the word set and the target trademark may be obtained based on a plurality of trademarks corresponding to the NICE field corresponding to the target trademark. Specifically, the degree of association between the word set and the target trademark may be obtained based on the number of times the corresponding word set appears in a plurality of trademarks corresponding to the NICE field. As an embodiment, the higher the number of times the corresponding word set appears in a plurality of trademarks corresponding to the NICE field, the higher the degree of association between the word set and the target trademark.

In step S340 , the server 10 may sort the plurality of words in order of importance based on the importance scores of each of the plurality of words.

Furthermore, the server 10 may group a plurality of words into n groups. In this case, a method of grouping a plurality of words may be various. In one embodiment, the n groups may be grouped to include the same number of words. As another embodiment, the n groups may be grouped according to a preset importance score. For convenience of explanation, it is hereinafter assumed that the importance score of the word included in the kth group is greater than the importance score of the word included in the k+1th group.

5 is a flowchart illustrating a method of obtaining a first word and a second word according to an embodiment of the present invention.

In step S410 , the server 10 may classify the plurality of words included in the target trademark into n preset groups based on the importance score.

In step S420, the server 10 may classify the plurality of words included in the similar trademark into n preset groups based on the importance score.

In step S430 , when the first word is included in a first group among n groups of the target trademark, the server 10 may obtain the second word as all words included in the similar trademark.

In step S440, when the first word is included in the kth group among the n groups of the target trademark, the server 10 assigns the second word to the first to (() n-k+1) can be obtained with words included in the group. In this case, n may be a natural number greater than or equal to 1, and k may be a natural number less than or equal to n.

For example, it may be assumed that the target trademark and the similar trademark are classified into three groups. At this time, when the first word is included in the first group, the server 10 compares the first word with all the words included in the first to third groups (that is, all similar trademarks) of the similar trademark literature to obtain the evaluation result. can be obtained Also, when the first word is included in the second group, the server 10 may obtain an evaluation result by comparing the first word with all words included in the first group and the second group of the similar trademark. Also, when the first word is included in the third group, the server 10 may obtain an evaluation result by comparing the first word with all words included in the first group of similar trademarks.

That is, the server 10 obtains evaluation results only for words determined to be important words according to the importance of the words, thereby preventing an evaluation result acquisition process for unnecessary or less important words, thereby reducing the amount of calculation. Specifically, since the first group of target trademarks is a high-importance word, it is compared with all similar trademarks, and since the third group of target trademarks is a low-importance word, it is compared with the first group of high-importance words among similar trademarks, so that the server ( 10), the comparison between the third group of (less important) target trademarks and the third group of (less important) similar trademarks can be omitted.

Meanwhile, according to various embodiments of the present invention, the server 10 may determine whether to group the target trademark and the similar trademark into how many groups. Specifically, the server 10 may determine the value of n based on the current GPU occupancy of the server 10 and required accuracy. When the GPU occupancy is high or the required accuracy is low, the server 10 may increase the value of n, and when the GPU occupancy is low or the required accuracy is high, the server 10 may decrease the value of n. For example, it may be assumed that the target trademark and the similar trademark are composed of 6 words, and n is 3 and n is 6. If n is 6, server 10 compares 6+5+4+3+2+1=21 words, if n is 3 server 10 compares 6+6+4+4+2+ 2 = 24 words will be compared. That is, as the value of n increases, the number of words to be compared in the target trademark and similar trademark decreases, and as the value of n decreases, the number of words to be compared in the target trademark and similar trademark increases. When the amount of computation (GPU occupancy) that can be processed is small or the required accuracy is low, the server 10 may increase the value of n.

Meanwhile, in the above-described embodiment, the case in which the target trademark and the similar trademark are grouped in the same number of n has been described, but the present invention is not limited thereto. For example, it goes without saying that the target trademark and the similar trademark may be grouped in different numbers.

In step S510 , the server 10 may obtain a similarity score and a dissimilarity score of the first word and the second word, respectively.

Specifically, the similarity score is an index for determining whether the first word and the second word are related words, and the dissimilarity score is an index for determining whether the first word and the second word are unrelated words.

In an embodiment, the similarity score and dissimilarity score may be obtained by acquiring semantic information of a set of words included in the first word and the second word, and determining whether the acquired semantic information is semantic information about the same subject. can

In another embodiment, the server 10 may determine the word structure of the first word and the second word and use it to calculate a similarity score and a dissimilarity score. For example, the server 10 may increase the similarity score when the word structures of the first word and the second word are similar, and increase the dissimilarity score when the word structures of the first word and the second word are dissimilar. .

In step S520 , when the dissimilarity score is equal to or greater than a preset score, the server 10 may determine that the first word and the second word are unrelated words.

In step S530, when the similarity score is equal to or greater than a preset score, the server 10 may acquire at least one word set not included in the second word among the word sets included in the first word.

That is, even if the first word and the second word are related words, the first word and the second word are not words having the same meaning. Accordingly, the server 10 may determine whether the first word and the second word are the same word or a word for the same subject or a word with a difference.

In an embodiment, the server 10 may determine at least one word set not included in the second word among the word sets included in the first word in order to determine the difference between the first word and the second word.

In another embodiment, the server 10 may determine the word structure of the first word and the second word.

In step S540, the server 10 obtains at least an importance score for each of the at least one word set not included in the second word, and a word set equal to or greater than a preset importance score among the obtained at least one importance score is selected. It can be determined whether it exists or not.

In step S550 , the server 10 may determine the first word and the second word as a matching word when there is no word set equal to or greater than the preset importance score.

In step S560 , the server 10 may determine the first word and the second word as mismatched words when there is a word set equal to or greater than the preset importance score.

That is, even when at least one word set of the first word is not included in the second word, there may exist a case where a word set having a low importance score is not related to the similarity between the first word and the second word. Accordingly, the server 10 obtains a word set when the importance score of at least one word set not included in the second word is equal to or greater than a preset importance score, and the first word and the second word are determined based on the obtained word set. It may be determined whether the word is a match or a mismatch word.

7 is an exemplary diagram for explaining a process of deriving an evaluation result using the first deep learning and the second deep learning according to an embodiment of the present invention.

As shown in FIG. 7 , the first deep learning may receive a target trademark as an input value and obtain a similar trademark for the target trademark. At this time, of course, there may be a plurality of similar trademark documents for the target trademark, and if there are a plurality of similar trademarks, the server 10 may sort the plurality of similar trademarks based on the association with the target trademark. .

The first database is a configuration for storing all trademarks. Specifically, the first database obtains a word set of each of the entire trademarks through morphological analysis of each of the entire trademarks, obtains a thesaurus for the obtained word set, and obtains an importance score for each of the obtained word sets Thus, the obtained importance score may be matched to a word set and stored.

The first deep learning may acquire a similar trademark corresponding to the target trademark among all trademarks stored in the first database.

Meanwhile, when the target trademark and similar trademarks are obtained through the first deep learning, the server 10 may extract a first word from among the target trademarks and extract a second word from among the similar trademarks. The first word and the second word may be obtained through the methods of FIGS. 4 and 5 described above.

Furthermore, in the second deep learning, the first word and the second word may be input as input values to obtain evaluation results for the first word and the second word. In this case, the evaluation result may be a result of whether the first word and the second word are a match word, a mismatch word, or an unrelated word.

The second database may store trademarks and prior research reports matched to the trademarks. In this case, the prior research report may include information on similar trademarks determined to be similar to trademark literature and specific words of similar trademarks corresponding to specific words of trademarks.

The server 10 may acquire the target trademark. As an embodiment, the server 10 may obtain a job invention from the electronic device 20 as a target trademark.

In step S605, the server 10 may obtain a word set by analyzing the morpheme of the target trademark.

In an embodiment, the server 10 may perform morpheme analysis of a trademark using a Mecab morpheme analyzer. However, the present invention is not limited thereto, and in some cases, various morpheme analyzers such as Okt, Komoran, Hannanum, and Kkma morpheme analyzers may be used. Furthermore, it goes without saying that the server 10 may use various morpheme analyzers according to the language used for the brand to be analyzed.

In step S610 , the server 10 may determine an erroneous word set among the plurality of acquired word sets.

In an embodiment, when the plurality of acquired word sets do not satisfy a preset condition, the server 10 may determine the corresponding word as an erroneous word. In step S615, the server 10 may correct the erroneous word set.

In operation S620 , the server 10 may acquire at least one set of compound noun phrases based on the degree of association between the plurality of word sets.

In one embodiment, the server 10 obtains a word included in the target trademark, classifies the text included in the target trademark by word unit, performs morphological analysis on each classified word, and converts the word into syllable units It is possible to segment, match a part-of-speech tag to the segmented syllable, and obtain a morpheme included in a word based on the segmented syllable.

In operation S625 , the server 10 may match semantic information for each of the obtained plurality of word sets.

In this case, the semantic information may mean an intent for a word set. The server 10 may perform natural language processing using deep learning to obtain semantic information about the word set.

Specifically, deep learning includes a natural language understanding unit, and the natural language understanding unit can grasp the intent of an entity and a word set included in a word based on the word analysis result, and further, the natural language understanding unit can determine the structure of a word And it is possible to interpret a word through analysis of major components, and perform word analysis using statistics/analysis, etc. FIG.

In an embodiment, the server 10 may obtain semantic information about the apple by analyzing a word including 'apple'. For example, it may be assumed that the word set obtained through the word grammar is “apple.” In this case, semantic information about apple may be semantic information indicating one kind of fruit as a noun, but other people as a verb. It may also be semantic information indicating that the user is wrong, and the server 10 may match semantic information determined to be suitable for a word among a plurality of semantic information on 'apple' with 'apple'.

In step S630 , the server 10 may obtain an accuracy score for each of the plurality of word sets based on the matched semantic information and the trademark.

In this case, although semantic information about the word set acquired through one word is acquired in step S625, the semantic information acquired in step S625 may be inaccurate. Accordingly, the server 10 may acquire the accuracy score of the semantic information matched to the word set based on the entire trademark including the word.

Accordingly, the server 10 may obtain an accuracy score for the semantic information of the word set obtained through word analysis based on the semantic information about the same word set found throughout the trademark.

For example, the semantic information about 'apple' obtained by analyzing a word containing 'apple' is information about an apple representing a fruit, but the semantic information of 'apple' that is searched for all trademarks including the word is different. In the case of information about telling the wrong person, the server 10 may set the pre-matched semantic information low.

In step S635 , the server 10 may determine a word set having an accuracy score equal to or less than a preset value as an erroneous word set.

That is, when the accuracy score is less than or equal to a preset value, the server 10 may determine that the semantic information matched with the corresponding word set is erroneously matched, and obtain the word set as the erroneous word set.

In step S640, the server 10 may obtain a plurality of semantic information for the erroneous word. Specifically, the server 10 may obtain a plurality of semantic information of the erroneous word from a plurality of trademarks.

In step S645 , the server 10 may acquire a plurality of weights for each of the acquired semantic information.

In an embodiment, the server 10 may obtain a weight based on a trademark including an erroneous word. Specifically, the server 10 may obtain at least one word including an erroneous word from the trademark, and obtain semantic information about the same word as the erroneous word included in the obtained at least one word. The server 10 may acquire weights for a plurality of semantic information based on semantic information about at least one word identical to the erroneous word.

In step S650 , the server 10 may obtain a weight having the highest degree of association with a brand among a plurality of weights.

In step S655 , the server 10 may match semantic information corresponding to a weight having the highest degree of association with a trademark to the erroneous word set. That is, the server 10 may determine semantic information having the greatest weight among a plurality of semantic information on the erroneous word as semantic information on the erroneous word and correct the erroneous word.

Meanwhile, according to various embodiments of the present disclosure, it goes without saying that the erroneous word set may be a case in which incorrect semantic information is matched with the word set, as well as a word set that is erroneously parsed due to an error in the morpheme analysis process. In an embodiment, when the word for obtaining the word set is “Natural language processing using machine learning is performed,” the server 10 may obtain machine, learning, use, natural language, and processing as a word set.

In this case, due to an error in morpheme analysis, a case of acquiring a word set such as machine, learning, using eul, natural language, and processing may occur. In this case, the server 10 may determine 'use E' as an error word set and correct it.

Specifically, after obtaining semantic information for each of the acquired word sets, the server 10 may determine whether the word is an erroneous word based on the degree of association between the semantic information and the trademark. According to an embodiment, when the acquired word set is found in a trademark including a word set with a frequency greater than or equal to a preset frequency, the server 10 may determine that the corresponding word set is an error-free word set. In another embodiment, when a word set is found with a preset frequency or more in all trademarks, the server 10 may determine that the word set is not an erroneous word set by determining that the word set is a commonly used word set. In another embodiment, when the server 10 does not find semantic information about the word set, the server 10 may determine the corresponding word set as an erroneous word set. In another embodiment, when the semantic information matched to the word set is different from a trademark including the word set, the server 10 may determine the corresponding word set as an erroneous word set.

Specifically, the server 10 may acquire a set of words included in the entire trademark, match and store semantic information for each of the acquired word sets. The server 10 may obtain words of a plurality of word sets having relevant semantic information by modeling the matched word set and semantic information. The server 10 may acquire word set words in which the number of word sets included in the word is less than or equal to a preset number among the plurality of word set words. In this case, the preset number does not mean the number of a plurality of identical word sets used in a trademark, but may mean the number of a plurality of word sets having different shapes. For example, even if the word "using E" is searched for multiple times in a trademark, "Using E" can be determined as one set of words because it is the same set of words. The server 10 may determine the word set included in the word obtained in the above step as the erroneous word set.

However, the present invention is not limited thereto, and it goes without saying that the server 10 may determine, among the plurality of word set words, a word set included in a word having a distance from another word equal to or greater than a preset distance as an erroneous word set. Specifically, the server 10 may obtain the center points of a plurality of words, and determine the distance between the words based on the obtained center points. For example, the obtained word set word may be a first word set word to a fourth word set word. In this case, the server 10 determines the distance between the first word set word and the second word set word, the distance between the first word set word and the third word set word, the first word set word and the fourth word set word When distances of are obtained, and all of the three obtained distances are equal to or greater than a preset distance, the first word set word may be determined as an erroneous word set.

In this case, the compound noun phrase set candidate may include order information and spacing information of word sets included in the compound noun phrase set candidate.

In an embodiment, the compound noun phrase set is determined as a combination of two or more adjacent word sets, but in the present invention, a combination of a word set in which words are spaced apart may also be obtained as a compound noun phrase set. For example, when the server 10 obtains a compound noun phrase set from a word including "a first device among a plurality of devices, a second device among a plurality of devices, and a third device among a plurality of devices", a plurality of The first device of the device, the second device of the plurality of devices, and the third device of the plurality of devices may be obtained as independent compound noun phrase sets, but "(spaced word 1) device of multiple devices" can be obtained as a single compound noun phrase set. Of course, it can also be obtained with In this case, the compound noun phrase set may include spacing information. In the above-described embodiment, the spacing information may be information that one word is included between the word “middle” and the word “device”. Of course, according to various embodiments, a plurality of words may be included between words.

In step S710 , the server 10 may obtain a word set that is a target for calculating the importance score from the target brand. Needless to say, not only the target trademark but also the importance score of the word set included in the similar trademark may be generated by the method described below.

In step S720, the server 10 configures the first detailed importance of the word set in the entire trademark, the second detailed importance of the word set in the trademark classification information corresponding to the brand name field information of the target trademark, and the word set in the entire trademark. It is possible to calculate one or more detailed importance among the third detailed importance of the search mark including .

In operation S730 , the server 10 may obtain an importance score of the word set based on one or more of the first detailed importance, the second detailed importance, and the third detailed importance.

In an embodiment, the server 10 may calculate the first detailed importance based on the first appearance rate and the second appearance rate. In this case, the first appearance ratio means the number of appearances of the word set for which the importance score is to be obtained in the entire trademark compared to the total number of word sets of the entire trademark, and the second appearance ratio is the appearance in which the word set appears among the words of the entire trademark. The number of words may be counted, and may mean the total number of words of the entire brand compared to the number of words in which the word set appears among the words of the entire brand.

Here, the trademark classification information is a code capable of classifying a trademark according to the brand name field, and may be NICE.

Here, W ₂ is the second detailed importance, NICEw is the number of occurrences of the word set in the trademark classification information, NICEW is the total number of word sets in the trademark classification information, and NICEs is the occurrence of the word set among all the words of the trademark. is the number of words, NICES is the total number of words in the entire brand, and a2 is the adjustment constant of the fourth appearance rate.

In another embodiment, the server 10 searches for a search trademark including a set of words to be calculated of an importance score among all trademarks, and calculates an influence value of each search trademark based on reference information of each search trademark and a third detailed importance level may be obtained based on the calculated influence value.

Specifically, the server 10 may calculate an influence value based on the number of items of at least one of the applicant, inventor, and right holder, which are reference information of the search trademark, the same as other trademarks, and the number of citations and citations that are reference information. .

That is, the server 10 may calculate the degree to which the search trademark is related to other trademarks as an influence value. For example, when a search trademark is cited from several trademarks, the server 10 may determine that the search trademark has an influence on the corresponding trademark, and may calculate an influence value of the search trademark.

Furthermore, the server 10 may calculate an average of the calculated influence values for each of the search marks, and calculate the calculated average as the third detailed importance level.

10 is a block diagram of an apparatus according to an embodiment.

The processor 102 may include one or more cores (not shown) and a graphic processing unit (not shown) and/or a connection path (eg, a bus, etc.) for transmitting and receiving signals to and from other components. .

The processor 102 according to one embodiment performs the method described with reference to FIGS. 1 to 9 by executing one or more instructions stored in the memory 104 .

For example, the processor 102 obtains new training data by executing one or more instructions stored in the memory, and performs a test on the acquired new training data using the learned model, and the test result, labeling Extracting the first learning data in which the obtained information is obtained with an accuracy greater than or equal to a predetermined first reference value, deleting the extracted first learning data from the new learning data, and removing the new learning data from which the extracted learning data is deleted It is possible to retrain the learned model using

On the other hand, the processor 102 is a RAM (Random Access Memory, not shown) and ROM (Read-Only Memory: ROM) for temporarily and / or permanently storing a signal (or data) processed inside the processor 102. , not shown) may be further included. In addition, the processor 102 may be implemented in the form of a system on chip (SoC) including at least one of a graphic processing unit, a RAM, and a ROM.

The memory 104 may store programs (one or more instructions) for processing and controlling the processor 102 . Programs stored in the memory 104 may be divided into a plurality of modules according to functions.

The steps of a method or algorithm described in relation to an embodiment of the present invention may be implemented directly in hardware, as a software module executed by hardware, or by a combination thereof. A software module may include random access memory (RAM), read only memory (ROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, hard disk, removable disk, CD-ROM, or It may reside in any form of computer-readable recording medium well known in the field of brand names to which the present invention pertains.

The components of the present invention may be implemented as a program (or application) to be executed in combination with a computer, which is hardware, and stored in a medium. Components of the present invention may be implemented as software programming or software components, and similarly, embodiments may include various algorithms implemented as data structures, processes, routines, or combinations of other programming constructs, including C, C++ , Java, assembler, etc. may be implemented in a programming or scripting language. Functional aspects may be implemented in an algorithm running on one or more processors.

In the above, the embodiments of the present invention have been described with reference to the accompanying drawings, but those skilled in the art to which the present invention pertains know that the present invention can be embodied in other specific forms without changing its technical spirit or essential features. you will be able to understand Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive.

10 : Server
20: electronic device

Claims (3)

In the method of determining the similarity of a trademark,
obtaining, by the server, a similar trademark by inputting the target trademark into the first deep learning;
acquiring, by the server, a plurality of words included in the target trademark, and acquiring a plurality of words included in the similar trademark;
obtaining, by the server, a first word among a plurality of words included in the target trademark, and obtaining a second word among a plurality of words included in the similar trademark;
inputting, by the server, the first word and the second word into a second deep learning step to obtain evaluation results for the first word and the second word; A method for determining the degree of similarity of a trademark, comprising:
According to claim 1,
The steps of obtaining the similar trademark include:
obtaining a plurality of word sets included in the target trademark;
obtaining a plurality of target brand words by modeling the obtained plurality of word sets;
Acquire a plurality of midpoints for each of the plurality of target brand words, and determine the location of the target trademark based on a plurality of midpoints for each of the plurality of target brand words and the number of word sets included in the target brand word to do;
obtaining a plurality of word sets included in a trademark, and modeling the obtained plurality of word sets to obtain a plurality of trademark words;
obtaining a plurality of midpoints for each of the plurality of trademark words, and determining the position of the trademark based on the plurality of midpoints for each of the plurality of trademark words and the number of word sets included in the trademark words; and
determining the trademark as the similar trademark when the location of the target trademark and the location of the trademark are within a preset distance; A method for determining the degree of similarity of a trademark comprising a. 3. The method of claim 2,
The step of obtaining evaluation results for the first word and the second word includes:
obtaining a similarity score and a dissimilarity score of the first word and the second word, respectively;
determining that the first word and the second word are unrelated words when the dissimilarity score is equal to or greater than a preset score;
obtaining at least one word set not included in the second word among the word sets included in the first word when the similarity score is equal to or greater than a preset score;
obtaining at least an importance score for each of at least one word set not included in the second word, and determining whether a word set equal to or greater than a preset importance score exists among the at least one obtained importance score;
determining the first word and the second word as matching words when there is no word set equal to or greater than a preset importance score; and
determining the first word and the second word as mismatched words when there is a set of words equal to or greater than the preset importance score; A method for determining the degree of similarity of a trademark comprising a.

Images