KR20050103836A - Method for trademark search including similarity decision and trademark information system using the same - Google Patents

Abstract

The present invention relates to a system for performing a search including the similarity of trademarks on the Internet, and to determine the possibility of registration of the trademarks, and more specifically, to quantify the similarity of trademarks and the degree of similarity using language engineering technology. The present invention relates to a method and a system for checking the possibility of registration before trademark application by checking in advance whether the trademark registration requirement is satisfied, including preparing a report and presenting it to the applicant.

Description

Method for trademark search including similarity decision and trademark information system using the same}

The present invention relates to a trademark retrieval method on a network and a trademark information system using the method, and more particularly, to a trademark retrieval method including a trademark similarity determination process and a trademark information system to which the method is applied.

The prior art related to the present invention is as follows.

First, trademarks, service marks, and other marks (collectively, the term 'trademarks') on the Internet, namely, trademarks, service marks, business marks, collective marks, and proof marks are classified by law, but unless otherwise stated, description For the sake of convenience, there is an application system for trademarks. However, these systems are only online processes of the trademark application process on-line. That is, when a visitor enters a trademark and selects a designated product (collectively referred to as a designated product including 'designated service industry' for technical convenience), the applicant prepares an application by electronic document method and uses the electronic application system of the KIPO. It is on behalf of. This is simply an online application of trademarks on-line, and lacks a preliminary judgment on the most important trademark registration requirements in the trademark registration application process. It does not provide any information at all.

Second, there is a program that automatically generates a brand during the brand naming process, but it is not only at the level of randomly generating a brand using a random function, but also whether there is a possibility of registration or whether another person uses the same or similar brand. It didn't provide any dictionary information, so I had to search for it manually.

Third, when searching for similar trademarks in the trademark search process, similar trademarks are uniformly generated by a computer program with a similar judgment process, but no quantified information or categorized information about how similar to the target trademark is provided at all. Or there is no basis for that judgment. In other words, not only the subjective judgment is entirely dependent on the trademark similarity judgment process, but the evidence data on similarity cannot be systematically provided by analyzing language engineering or language science.

The technical problem to be solved by the present invention is as follows.

First, it is about the linguistic engineering process and various systems for determining the similarity of trademarks. Specifically, a trademark processing process including a product similarity determination engine 200-1, a trademark similarity determination engine 100-1, an engine for extracting a part without identification, and a language processing system when two or more languages are involved, Trademark DB, reporting system.

Secondly, it is an all-in-one process for determining the possibility of registration of a trademark and a system therefor. In more detail, it relates to a registerability determination engine 400-1 for processing various registration requirements, various DBs, a reporting system, and a registerability processing processor.

Definition of Terms

The network 1000 refers to a wired or wireless communication network.

Offline refers to the state of not being connected on the network 1000 as a (relative) opposite concept of online.

Online comprehensively defines the environment of all types of network 1000 terminals (including mobile, geographic information systems, satellite receivers such as transmission and reception devices using communication satellites) that can be connected to the network 1000. .

Offline media are physically shaped media, which include analog recording media such as video, tape, and paper media including books, floppy disks (FDDs), CDs (CD-ROMs), Used as digital data storage media such as digital video discs (DVD), magneto-optical discs (MOD), hard disks (HDDs), RAM, ROMs, packs, photo masks, etc. Refers to any device.

Synchronization means that computer users interact with each other online or with online and offline digital content so that the online and offline content can be connected, combined, merged, and exchanged. A method provided to a computer user through a process of reorganization or transfer.

A computer user refers to a user who uses a computer. In the present invention, a computer refers to a PC, a PDA, a notebook, a palm top, an e-book, a work station, a mini Central processing unit (CPU) such as computer (mini computer), super computer, cellular phone (mobile phone), pager (beeper), game machine (e.g., Sony Play Station, etc.), digital TV Or a separate device attached to a built-in central processing unit to perform a function as a physical communication terminal device, a device including a numerical processing unit (NPU), whether wired or wireless. ) Is an arbitrary machine with the function to send and receive data by

The internal memory device is a concept that comprehensively defines the media that can be read and written among the internal storage device and the auxiliary storage device. Volatile and nonvolatile memory devices such as RAM or ROM, and optical recording devices such as CD-RW are devices that can be read and written and mounted in a computer.

Trademark: In the present invention, a trademark mainly includes a trademark and a service mark, but is broadly defined in commercial law such as any identification mark and trade name defined in a law related to a mark such as a trademark, a service mark, a collective mark, a business mark, etc. Identification mark, identification mark as defined by the Anti-Competition Law, identification mark, model name of the product, any cover and website address (including sub-site address) used for the purpose of identifying other products, icons, images, symbols ), Sound labels, slogans, gestures, specific scenes in advertising, specific sounds, CM songs, CFs, etc. Contents used for any purpose of identification or having the function of identification, or any combination thereof We say thing corresponding to two or more. The above law refers to the laws or treaties of the Republic of Korea and foreign countries related to the emblem. Among these marks, "trademarks" may be substituted for "trademarks" unless there is a special mention for the convenience of reality and description, which is handled by the socially significant weight of trademarks.

Designated goods: In the present invention, designated goods are designated goods or designated services business as their main targets, but broadly related laws of Korea and foreign countries related to marks of designated goods, designated services, business, business, etc. Or one or two or more of those designated for the purposes of each mark as provided for in the Treaty, the name of the goods or services in the field of merchandise, the names of the goods or services that are socially accepted, or the type of business. The term "designated product" may be substituted for the term "designated product" unless otherwise specified for the convenience of the point and description described in the Korean Trademark Act.

That is, when a service mark is involved, a trademark is a service mark, a product or a designated product is a service industry or a designated service industry, and a trademark DB is a service mark DB (except when the trademark DB includes a service industry DB.), And the name of the product is a name of a service business. The structure of mapping a service mark to a trademark related position is taken. This is for convenience of description, and the purpose of this article is “Regulations concerning Trademarks in this Act, except as specifically provided for in this Act concerning service marks, collective marks and business marks” in Article 2, paragraph 3 of the Korean Trademark Act. Is applied ”.

Applicant, etc .: Any person involved in judging the similarity and registration possibility of a trademark in all areas to which the spirit of the present invention is applied is collectively referred to as an "applicant or applicant", regardless of whether or not the actual application. However, when the other party is specified or specified in relation to the applicant / applicant in the specification, it refers to a person excluding the other party.

Trademark DB: The trademark DB refers to a DB having any information related to a trademark in the present invention. Trademark DB is a database containing domestic or foreign registered, filed, extinguished, unapplied trademarks and all information related to the trademarks.All trademarks including the right of trademark rights and any product using the trademarks. Information on the subject, such as the applicant, any information on the trademark, bibliographic matters on the trademark, any proceedings in relation to the Patent Office, the Court, etc. Refers to a DB containing some or all of the processed internal or external information, or any information about the processing of the trademark (2-3). Any DB for a registered or filed trademark of the JPO may be a component of the trademark DB in the present invention. Broadly, the trademark DB may include a DB containing information about the product.

Configuration of the entire system

Overall composition

The configuration of the entire system according to the spirit of the present invention will be described with reference to FIG.

Since the present invention supports the similarity determination and registration possibility determination process related to a trademark on a network or a system interworking with the network, each computer is connected to the wired / wireless network 1000. The system server 1-1 is at the center of performing a specific function such as similarity judgment, and according to the function, the trademark similarity judgment system 100, the product similarity judgment system 200, the similarity judgment system 300, and the possibility of registration It is divided into a judgment system 400. Of course, one system server may perform two or more functions simultaneously. Each function system server has an engine and a DB (trademark DB 700, member DB 710, and various other DBs 750) for performing the function, and a management system for managing the internal or external of the system ( 1100). In addition, the system server 1-1 is connected to the applicant, member or customer (or their computer) 800, which may be an individual but may be a patent law firm 810 or a legal entity or other legal entity. . Applicants, members, and customers may also be connected to portable computer 820. The system server 1-1 may include various external servers 1220 such as an external agency server 840, a patent office server 1200, a foreign patent office server 1210, domestic and overseas customer company servers 830 and 840, and / or domestic and foreign customs office servers. ) May be connected to the Each of these servers may include various data DBs 750 for performing the functions of the server.

Management system (1100)

Introduction of a management system is essential in the course of carrying out some or all of the inventive idea. The management system is composed of processes similar to those of a free and paid service provider on a normal network 1000 that provides a web service to a member or a visitor. The functions performed by the management system include member-related functions such as membership registration, account management, and all procedures related to terms and conditions, billing or payment-related functions that can be combined in each process of the present invention, and storage and backup of various materials ( The management system is in charge of system integrated management functions such as backup) function, communication with external servers, and any additional functions for security and web services.

In addition, in order to access each system, 1) a method of accessing and using a virtual space that provides individual services through membership registration and membership authentication process (or including some or all of the services without membership registration), 2) there may be a method of embedding some or all of the functions of each system or DB into the computer of each user 800 through an online / offline medium, and using the functions of each system in conjunction with a network. The former is usually in the form of a web service, the latter can be in the form of a web interworking solution. Detailed configurations thereof are included in the general technical field, and thus the description thereof is omitted.

DB used in the present invention

DB is widely used in the present invention. These databases generally include trademark DB, product DB, data DB, DB for each system, other DB related to this system, and / or external DB.

The trademark DB includes the contents described in the definition of terminology, in particular, the registered trademark DB 710-1 (this trademark DB includes any matters relating to the registered trademarks—trademarks, designated products, owners and rights relations, applicants, Any information about the registered trademark, such as information on the agent, any bibliographic information, judgment relations, or the progress of the application, may be the contents of the registered trademark DB, which is the same for the trademark DB and the rejected trademark DB in the application. ), Pending trademark DB (710-2), rejected trademark DB (710-3), similar trademark DB (710-4), famous trademark DB (710-5), publication DB (710-6), and other patent office materials The main contents may be DB 710-7 and / or related DB 710-8. As a source, a trademark DB may be derived from a patent office and may be unrelated to the patent office (for example, a customer's trademark DB, an unregistered trademark DB, etc.).

The merchandise DB includes the product classification / classification DB 720-1 of the Patent Office (this may include classification and NICE classification), foreign product classification DB 720-2, HS code DB 720 -3), a product dictionary and other product classification DB 720-4, a similar product DB 720-5, and the like. When each product classification system is stored, it becomes a DB by itself. In addition, there is a matching engine 720-6 for establishing a relationship between each product classification system. In other words, the Korean Patent Office has data that has a relationship between classification and new classification. An example of a matching engine is to automatically process this relationship.

The data DB includes trademark processing-related DB (730-1) such as dictionaries, judgment / litigation / precedent data DB (730-2), orthopedic DB (730-3) (730-3), and non-standard DB (730-4). 730-4).

Each system DB 745 has DBs for performing the purpose of the system for each system. The details of these databases are in the description of each system. There may be a similarity determination system related DB 745-1, a registration possibility determination system related DB 745-2, and the like.

The external DB 740 includes a customs office registered trademark DB 740-1, an overseas trademark DB 740-2, a registered trademark of each country, a pending trademark, other trademark DB or related DB 740-3, and other external DB. (740-4) and the like.

Other system related DBs (746) include member DB (746-1), competitor DB (746-2), each registration requirement determination rule DB (746-3), each registration requirement data DB (746-4), identification power Element DB 746-5 may be missing.

Similarity Determination System (300 / 1-1)

Concept and Necessity

Similarity determination system 300 / 1-1 refers to any system for achieving the purpose of determining the similarity of trademarks defined in the present invention. The similarity determination system 300 / 1-1 can be broadly divided into a product similarity determination system 200 / 1-1 and a product similarity determination system 200 / 1-1. Therefore, the description of the similarity determination system 300 / 1-1 will be described by dividing the trademark similarity determination system 100 / 1-1 and the product similarity determination system 200 / 1-1.

This is consistent with the practice that, even if the same or similar trademarks are different, if the designated product is different, the trademarks are not entirely similar unless there is a misunderstanding of the source on the basis of the general consumer. However, in the present specification, unless otherwise stated, such as the product similarity judgment (5-2), the word similarity judgment (5-1) is based on the product similarity judgment (5-2).

The trademark similarity determination system 100 / 1-1 is composed of 1) hardware and network 1000, 2) software and various DBs.

An external third party may access the trademark similarity determination (5-1) determination system. In this case, the external and registration possibility determination system 400 / 1-1 may be configured to: 1) The trademark similarity determination system 100 / 1-1 may be used. Existing server, 2) at least one user (applicant, customer, member, etc.) terminal, and 3) a network 1000 system connecting them.

Determination of trademark similarity is essential for 1) determining the registration requirements of a trademark, 2) determining the scope of a trademark's rights, and the actions, procedures, requirements, and regulations that relate to, or are included in, sub-concepts thereof.

The criteria for judging trademark similarity may, in fact, be a variety of legal standards, which may be based on the internal or customary standards of each country's legislation or national judiciary (typically becoming an administrative or judicial body for trademarks). can be different. At present, the Korean Intellectual Property Office takes the attitude of judging the similarity by using the appearance, title, idea, etc. as the main judgment factors, and the Supreme Court also judges the similarity on its own basis. However, they are all relative standards, not absolute standards. Therefore, the similarity determination system 300 / 1-1 of the present invention aims to avoid absolute similarity and to derive the degree of similarity based on mathematical, statistical and linguistic models on how similar the target brands are.

Trademark Similarity Determination System (100 / 1-1)

The trademark similarity determination system 100 / 1-1 refers to a system for determining whether a target trademark is similar to other trademarks. The system includes 1) a trademark similarity determination engine 100-1 for determining trademark similarity, 2) a trademark similarity determination rule DB, and 3) a trademark input system for inputting a target trademark to the trademark similarity determination engine 100-1 (2). -1) essentially. In addition, the similarity determination result output system 300-4, the report preparation system 300-5, or the online electronic filing system 300-6 may be selectively combined with the trademark similarity determination system 100 / 1-1. . The same is true in the product similarity determination system 200 / 1-1, and the same in the similarity determination system 300 / 1-1 including both a trademark and a product.

Trademark Similarity Determination Engine (100-1)

The trademark similarity determination engine 100-1 is composed of various programs and a database DB for determining trademark similarity. The programs relate to theories and applications of AI theory, linguistic engineering, mathematical modeling techniques, and other computational methodologies for determining brand similarity.

The trademark similarity determination engine 100-1 is configured as a module for forming a trademark similarity determination rule and an application module for the trademark similarity determination rule 100-2, and is mutually related to the trademark similarity determination rule DB 100-2. Works.

Trademark Similarity Determination Rule (100-2) Formation Processor

Basic process of forming the trademark similarity judgment rule (100-2)

In order for the computer to automatically determine the similarity of trademarks, trademark similarity determination rule 100-2 is required. The trademark similarity judgment rule 100-2 may be of two kinds: a rule that recognizes trademark similarity and a rule that denies trademark similarity. This rule is largely formed through three processes.

First, it is an arbitrary setting step of the rule similarity judgment rule group. This is how a person arbitrarily sets a set of preliminary rules for determining trademark similarity. For example, a person may rule out “a” and “a” or similar trademark elements in a letter trademark. Each of these trademark similarity determination rules 100-2 has a unique identification number, and if applied in practice, each rule may be weighted.

Second, it is a step of judging how well the trademark similarity judgment rule 100-2 is applied through the trademark similarity determination data. In other words, for each trademark pair for which trademark similarity is recognized, 1) determine which trademark similarity recognition rules are applied well and which trademark similarity fraud rules are applied incorrectly, and 2) statistically process them to determine how often Or you can figure out whether it was applied correctly or incorrectly by weight. In addition, for each trademark pair whose trademark similarity is denied, 1) identify which trademark similarity fraud rules were applied and which trademark similarity recognition rules were applied incorrectly, and 2) statistically process them to determine how often Or you can figure out whether it was applied correctly or incorrectly by weight.

Materials for which trademark similarity is judged are called similar trademark DB 710-4, and the following are examples. 1) Trademarks that are legally similar (trademarks based on established domestic and international precedents, trademarks judged by the public office for the Patent Office or other trademarks), and 2) Internal or internal examination by the Patent Office or other administrative agencies in or outside Korea. Documents not publicly or privately judged for trademark similarity, such as domestic and foreign similar trademark casebooks made for reference purposes; 3) Vocabulary or domestic / local phonebooks, applications or registrations issued by the Korean Intellectual Property Office Some or all of the data arranged in a certain order (such as alphabetical order, alphabetical order, etc.) may be similar trademark DB 710-4. It is common for patent offices in each country to have a collection of similar trademarks. Referring to the data existing in the Korean Intellectual Property Office as an example. Similar trademarks or collections of similar vocabulary are, in practice, examples of helpful materials to determine trademark similarity in a mark. Although there is no legal binding force in the similarity judgment, it is an important measure for the judgment of trademark similarity and, in the present invention, these materials can be used as learning materials for the trademark similarity judgment program.

Third, a weight is determined for each of the trademark similarity determination rules 100-2. Naturally, each trademark similarity judgment rule 100-2 does not have the same weight. Therefore, weights indicating how well each trademark similarity determination rule 100-2 is applied need to be determined.

This can be briefly expressed as a mathematical expression as follows.

The degree of similarity = sum {the degree of application of the ith similarity judgment rule * the weight of this rule}

Formation of trademark similarity judgment rule (100-2) through the use of mass DB

The trademark similarity determination rule 100-2 may be extracted by using a large amount of DB for which trademark similarity is determined. Similar trademarks (including foreign counterparts), textual materials listed under a set of rules, such as national dictionaries or industry directories, are examples of large databases. In this database, trademarks that have been judged for similarity during the examination, judgment, and litigation process may be components.

Linguistic characteristics can be arbitrarily extracted through statistical methods in large databases that are judged for similarities such as similar trademarks. This can be achieved through conventional pattern extraction techniques in computational statistics. Through this pattern extraction technique, many meaningful patterns or rules can be found by the attributes of the input data even if they are not meaningful or mathematically meaningful. These found patterns or rules may form part of a similarity determination rule and may be weighted according to how often the pattern or rule is applied (ie, the degree of repeatability).

The process of forming the trademark similarity judgment rule 100-2 using this mass DB is as follows.

First, a step of inputting a large DB into the similarity decision rule forming module. In this case, the input may be input by making a similar trademark pair.

The second step is to find meaningful patterns or rules in the similarity decision rule forming module using computational statistical methods. Of course, in this case, a simple pattern or rule may be presented first, and then, how well the pattern or rule itself is applied, or a substructure pattern or rule of the pattern or rule may be found. In this process, the weight of the pattern or rule may be determined.

Third, structure the extracted pattern or rule. That is, the structure of the rule may be configured by upper sub category by classifying a pattern or a rule. This process can be done by humans, or they can select patterns or rules extracted against the practical and linguistic structure of similar judgment.

Fourthly, the extracted pattern or rule is reviewed for generality of its application through another similar trademark DB. In particular, it is possible to determine the degree of applicability of the pattern or rule through specific examples determined in the examination, judgment, litigation, and the like.

The above process can also be performed for a database of examples in which similarity is denied. What comes out of this process is a pattern or rule that denies similarity. This denying rule may also form part of the similarity determination rule DB.

Improvement of Trademark Similarity Determination Rule (100-2) Formation Module

The trademark similarity determination rule 100-2 forming module can be improved. The improvement is largely three axis.

The first is to change the weight of the trademark similarity determination rule (100-2) by additionally inputting new data determined trademark similarity, whether or not to apply each rule.

The second is to add or delete a new trademark similarity judgment rule 100-2. In the case of adding, the process of forming the trademark similarity determination rule 100-2 consisting of the above three steps may be followed. In the case of deletion, the weight of the deleted rule may be set to 0, or the deleted process may follow the process of forming the trademark similarity determination rule 100-2, which is composed of the above three steps.

Third, there may be a case where the trademark similarity determination rules 100-2 depend on the characteristics of each trademark. That is, each trademark may be grouped based on a specific attribute, and the weight of the trademark similarity determination rule 100-2 within the group may be determined. In this case, if the contrasting trademark belongs to that particular group, then that group belongs to that group in an independent, preferential, or weighted average (by weighting averaging general rule application and attribute rule application, etc.). The trademark similarity judgment (5-1) can be made by reflecting the weight.

Generalization, direct and indirect preparation of the trademark similarity judgment process

There are essentially two judgment processes for similar judgment.

One can essentially reduce the two trademarks (or similarity of n trademarks) to the contrast pairs of the two trademarks. The number of contrast pairs will be at most nC2, but if the problem is centered around one trademark, If you contrast the trademark, it is the process of judging the trademark similarity directly.) This process is defined as 'direct preparation process'.

The other is the process of extracting attributes from each trademark or matching the numerical values, and then determining the similarity of each trademark in question based on these attributes or numerical values. This is called an indirect contrast process.

The indirect preparation process can create and store a database that gives attributes or numerical values to a number of trademarks, and has the advantage of processing large amounts of trademarks or problematic DB data at once. The direct contrast process provides a direct contrast of 1: 1, so that the application of the comparison function to the contrast process of the two brands in question can be observed in more detail and directly. Each will be described below.

Direct preparation course

C is called a direct contrast element function vector, and C = C (T1, T2) = [C1 (T1, T2), C2 (T1, T2),... Cn (T1, T2)], where T1 and T2 are trademarks entered into the engine. C1, C2,... , Cn is the direct contrast factor function. The direct contrast function B = B (C) = SIGTC (Wi * Ci), where Wi is the weight. At this time, the direct contrast function vector C may be defined, where C = C (C) = [C1 (C1), C2 (C2),... , Cn (Cn)]. Through this process, a B value or a C value is determined to correspond to both trademarks T1 and T2, and the similarity of the trademarks can be determined by comparing these values (which may be a reference vector or the same) with the reference.

Indirect Preparation Process

X is called an indirect contrast element vector, where X = X (T) = [X1 (T), X2 (T),... , Xn (T)] and (where T is a trademark input to the engine). X1, X2,... , Xn becomes the indirect contrast factor function. The indirect contrast function is defined as Y = Y (X) = SIGTX (Wi * Xi), where Wi is the weight. At this time, the indirect contrast function vector Z may be defined, where Z = Z (X) = [W1 * Z1 (X1), W2 * Z2 (X2),... , Wn * Zn (Xn)]. Through this process, the Y value or the Z value (which may be a reference vector) is determined to correspond to the trademark T, and the similarity can be determined by comparing them.

Determination of weight

The weight (which can be a weight vector (if n is 2 or more), and each weight constituting the vector may be a weight function, is the same above) is then determined to prepare for each trademark in question or a contrasting trademark. Determine a function or contrast function vector. Therefore, the determination of weight is the key to similarity judgment. How to determine the weight is largely as follows.

First, there is a method of determining through a learning process by inputting a similar brand DB (710-4). Algorithms used in the learning process include artificial intelligence techniques (which refer to any technique that results in improved weights due to the input of more accurate data) and other algorithms consistent with the purposes of the present invention.

Second, there is a method of inputting a weight arbitrarily by a person.

Third, there is a method of using a combination of a learning process and a human input process. This is especially useful when determining the initial weights and applying the present invention in part to separate trademark areas. That is, the initial weight may be arbitrarily input, but the process of determining the weight may be improved by inputting various similar judgment cases occurring thereafter, and the weight may be determined at each moment. This is described in detail in the trademark similarity determination rule 100-2 forming module.

Direct Contrast Factor Function

In order to determine the similarity of the two trademarks in contrast, each trademark may be contrasted from various points of view, each of which is a direct contrast factor function.

Indirect Contrast Factor Function

When a property is extracted through an analysis process or a numerical value is mapped to a trademark, the extraction process or the numerical value matching process may be an indirect contrast factor function.

Rule DB

Once the trademark similarity judgment rule 100-2 is determined, the rule is stored and managed in a DB. Each rule may be weighted for that rule, and there may be rules that apply only under certain conditions or trademark attributes.

Processor applied to trademark similarity judgment rule 100-2

If there is a trademark similarity rule DB, the trademark similarity determination rule 100-2 may be applied to two or more target trademarks. The steps of its application are as follows.

First is the processing of the target trademark. Allow the trademark similarity determination rule DB 100-2 to be applied (including allowing other determination engines of the present invention to process) the target trademark. Processing, segmentation, analysis, transformation, localization or extracting of essential parts, etc. Rule DB (hereinafter referred to as “trademark processing (2-3)”, which is particularly important in the processing of combined trademarks) allows the rule DB to be applied. That is, the rule DB has a form of a function, which is to process the trademark into a form that can be input to the function.

Second is the application stage of each rule. Various trademark similarity determination rules 100-2 are applied to the target trademarks to determine whether the rules are applied, and the result of the determination is output.

Third, the weight is applied to the result of determining whether the rule is applied to finally determine whether or not the trademark similarity.

Trademark similarity can be mathematically expressed as C (T1, T2) = sum (Ri (T1, T2) * W (Ri)). C is a trademark similarity judgment 5-1 function, Ri is each trademark similarity judgment rule 100-2, and W (Ri) is a weight for the rule Ri. At this time, the rule Ri may be dependent on the specific attribute of the trademark group. In this case, the weight may be expressed as W {Ri (Gj)}, where G is a trademark group having a specific attribute, and W {Ri (Gj)} denotes the weight of Ri when the trademark corresponds to a specific j attribute. Say.

Application of trademark similarity judgment rule (100-2) and reporting evidence of trademark similarity judgment (5-1)

Evidence for the determination of trademark similarity needs to be presented and needs to be output to users or other parties. This output is referred to as the basis of trademark similarity judgment (5-1), which includes 1) whether each trademark similarity judgment rule 100-2 is applied, 2) the weight of the rule, 3) the basis of the rule's formation, 4) When various rules are applied, one or more of the statistical processing results in the case where each weight is applied. Performing this process results in the similarity determination result output system 300-4 or report generation system 300-5.

Input of target trademark

Concept and Necessity

Trademarks are defined differently in each country's legislation, and the Korean trademark law defines the elements of a trademark by combining colors with symbols, letters, figures, or a combination thereof or each of them. Among these components, the material of similar judgment in the idea of the present invention is 1) a symbol or a character that is imported into a domestic or foreign language or symbol expression system (the image is converted into an OCR (Optical Character Recognition) program). Computer-recognized characters.), 2) The data entered by the person judged by the target trademark (if there are “heavenly sky” characters drawn in the circle, the image of “天” or “heaven” entered by the person) Seeing an existing trademark and entering characters recognized from that trademark (such as an image containing TIFFANY, a textual representation of something under human recognition, such as TIFFANY, tiffany, or Tiffany) or a new picture Inputting the meaning of a figure into an input field of the notion of “new”, etc.), or 3) a member of a trademark that he / she intends to file or search for (Sound) "is such a representation that is part of the character input system determines the destination of the Trademarks oil processed by computer.

Processing (2-3) of target trademark

Depending on the nature of the target trademark and the attributes of the input trademark of the trademark similarity determination engine 100-1, processing may include the following.

First, it is the work of localization or language unification. For example, to determine the trademark similarity between “Tiffany” and “Tiffany”, the unification work of language (Hangulization refers to a special case of unification of language unification into Korean. Essentially the same as unifying). In this case, we need to localize the Tiffany. This localization work has been described in the section on English localization described later.

Of course, you can work without language unification. In other words, if there is Tiffany, it can be judged by itself, or it can be determined by changing to Hangul, or both can be performed at the same time. Of course, in the case of Korean and English (for example, a trademark consisting of Tiffany at the top and Tiffany at the bottom), both should be judged.

Second, processing each trademark so that the rule DB can be applied directly. For example, if the rule is an arrangement of consonants and vowels, an example would be to divide Tiffany or Tiffany's Korean characters into consonants and vowels in order.

Third, when a trademark is a combined trademark, it may be processed to extract only the distinguishing ones from each trademark.

Fourth, the processing (2-3) of the target trademark can be converted into a system that can be recognized by the computer.

Fifth, as a premise in each judgment processor, it is possible to first determine which language system the trademark component is in. There may be a combination of a Korean and a foreign language only in the Korean language. When a foreign language is included, an internalization process may be performed internally, or an applicant may request input of a Hangul pronunciation corresponding to the applicant. Of course, you can also proceed in a foreign language itself. In the case of a trademark in which both Korean and English are written in the upper and lower sides, considering that it is the position of the Korean Intellectual Property Office, the determination processor of the present invention may proceed separately from Korean and foreign languages. This is especially true when the meaning or pronunciation of a foreign language is not related to the written Hangul. For example, if there is a correlation, such as when there is a "tiffany" notation at the top of the "tiffany" at the top, it is possible to proceed with various decision processors for either or both. However, if there is no relevance or lack, such as the case of "fab" at the top and "SAMSUNG" at the bottom, it is necessary to proceed with various judgment processors for both.

The trademark DB of the Korean Intellectual Property Office is open to the public in the state of being localized or processed at a certain level.

Discrete Element DB (746-5)

Elements without discernment are not subject to comparison in trademark similarity judgment (5-1). Therefore, it is necessary to extract only the distinguishing elements in such a manner as to remove the non-identifying elements among the elements constituting the trademark. Hereinafter, a method of finding an element without discernment will be described.

Unidentifiable trademark component discovery process through mutual DB

Analyzing mutual phone book DB, business phone book DB, and registered company DB (hereinafter referred to as 'interrelated DB') where data on company or business companies are accumulated, have no common name or discernment You can find many syllables, words, words, and more. The non-identifiable element obtained through this process becomes a component of the DB with no discriminating element. This process is as follows.

First, retrieve each other from the related DB. At this time, although there may be one type of mutually related DB, there may be a plurality or a plurality of types of DB.

Second, they perform a mutual segment. Segment means to first determine whether a mutual name is a combination of two or more meaningful phrases, and in this case, the characters constituting the name are regarded as a combination of two or more phrases having minimum meanings, and extract each of the phrases. At this time, whether or not the meaning is meaningful is a variety of dictionaries such as Korean dictionaries, abbreviations such as abbreviation dictionaries, etc. Other idioms of general industry (eg, 'construction,' building ', etc., abbreviated as civil construction industry) Can be used as a reference to identify In addition, one mutual name may be capable of two or more types of segments. In this case, two or more types of segment pairs may be stored independently in a DB. When there is a special internal standard, DBs may be assigned to different segment pairs by assigning different weights to each segment pair. Can also be stored in

Third, the segmented DB can be sorted or counted to store the degree, frequency, and frequency of repetition of specific syllables, words, words, etc. independently or in groups. . In other words, if the mutual group is a construction related mutual, it is possible to find a mutual component that is often used in the construction related trade.

Fourth, the degree of repetition, frequency, number of times, and the like can be compared internally with a given criterion to identify that there is no discernment. The criteria may be given uniformly for the entire trade name, but may be given differently for each trade group.

Through this process, the non-identifiable part of the trademark can be extracted.

Industry-specific phonebooks can be used to extract indistinguishable elements from specific industries. Extracting non-identifiable elements is in principle the same as using a mutual directory DB. However, what can be found through the sector-specific telephone directory can be found in the trademark components that are not identifiable in a particular upper or lower sector or sector. For example, in the medical industry, "doctors" can identify information that is indistinguishable. In this case, when a correspondence relationship is established between the upper or lower industry type or the industry type group and the industry type, a corresponding relationship of 1: 1, n: 1, 1: n may be established with the designated service industry. This response process may identify trademark components that are indistinguishable for the designated service sector corresponding to a particular higher or lower sector or sector.

Identification process without identification through trademark DB

The process of analyzing the above non-identifiable trade names can also be applied to the trademark DB. At this time, the trademark may be processed in all of the premise trademark, or by each class, group designation product or other unit group. In addition, a trademark entered in the analysis may be a group having certain attributes in relation to a registered trademark, a pending trademark, or a trademark. Through this process, it is possible to identify elements that are not or lacking in identification by category, group, and designated product. The trademark DB input in this process may be not only a registered trademark DB 710-1 but also a DB containing information about a pending trademark DB 710-1 and a rejected trademark. In particular, the non-identifiable element extracted only through the rejected trademark DB may be given more weight than the non-identifiable element found through other DBs. In particular, a greater weight can be given to non-identifiable elements extracted from a database that has been rejected by applying the regulation on technical expressions (Article 6, Paragraph 1, No. 3 of the Trademark Act).

These non-identifying elements are stored in the non-identifying element DB. This DB may contain information about elements that are not discernible only when combined with designated goods, groups or classes.

Problem of Combo Label

Concept and material of problem

Trademarks that can be decomposed into two or more trademark components are called combined trademarks. Since a bond is usually a combination of a number of components, it is problematic to extract the strength and discernible portion of the bond. Since it is the precedent of a separate observer, unless it is unnatural to separate and identify it, it is important to separate it by the elements of the combined trademark. The following is a processor on how the system automatically separates when a binding label is given.

Lumbar extraction engine

The main part extraction engine refers to an engine that extracts the distinguishing part from the combined trademark, and has the following configuration and procedure.

Separation steps

When there is a trade mark, the first task is to separate it into its components. In the natural language processing technique, the character part of a given combined trademark is cut using the cutting rule. Morphological separation may be a criterion of the cutting rule. What this requires is a morphological analysis engine.

The process of separating trademark components through the stemming engine is as follows. First, the binding trademark is input to the stemming engine. In this case, when the combined trademark is spaced, 1) the spaced unit may be a separate element, or 2) the morpheme may be separated again within the spaced unit.

Second, when the morphemes are separated by each other or by other criteria, a plurality of cut sets may be generated. This cutting set is stored in association with the binding label. In the case of determining discrimination or similarity of the combined trademark later, the presence or absence of discrimination power can be judged by the individual cut parts constituting the set for each of the plurality of cut sets.

Elimination of non-identifiable parts

Commonly used stock companies, firms, unions, etc. (including foreign language counterparts) should be removed because they are not discernible. The DB for the lack of such discrimination power can utilize the extracted elements with high commonality through the analysis of the phone book DB. At this time, it can utilize the element DB without identification.

The verbs and adjectives in the dictionary or their modified forms (changes in verbs, derivatives, etc.) may in principle be indistinguishable, especially when applying the technical labeling rules. This is especially true when the other parts of the combined trademark are not discernible.

Truncated parts that are not in the nouns or the dictionary DB are not removed in principle, and the process proceeds to a later registration decision process. The cut part may be a common name or a common name in relation to the designated product, and the presence or absence of the discrimination power thereof is determined by the registration possibility determination system 400 / 1-1 described later. The same is true when geographical names are combined.

It is possible to determine the remaining part after removing the adjective letter from the trademark formed by combining the adjective letter.

Element DB without discernment

You can cut the phone book, website address, or other personal names and business names through the cutting rules, and then check the frequency of occurrence of each cutting element. At this time, the occurrence frequency is higher than a certain internal standard, it becomes an element without discrimination, and this stored becomes the element DB without discrimination. Normally, Co, www, com, etc. do not have discriminating power.

Associate a trademark DB (including both domestic and international applications and registered ones) with a designated product that uses that trademark and, on the basis of the classification unit, corps unit, or the designated product itself to which the designated product belongs; Can be analyzed. That is, by cutting each trademark of the trademark DB, the occurrence frequency of the specific cutting element can be examined for the specific designated product, the group to which the designated product belongs, the class to which the designated product belongs, or the whole product. At this time, it is considered that there is no discrimination when the cutting element appears at a frequency higher than or equal to a predetermined standard with respect to a specific designated product, a group to which the designated product belongs, a class belonging to the designated product, or the whole product. The non-identifiable element is stored in a relational state with respect to the designated product, the group to which the designated product belongs, the class to which the designated product belongs, or the whole product, thereby forming an element DB without discrimination. “Home” / ”home” etc. may not be discernible for confectionery and household items.

The fact that a person's attention is not usually discriminated by the public's attention can also be a factorless DB. In addition, the fact that discrimination is denied through precedent or judgment may be added to the element DB without discrimination before or after.

Lumbar extract

The non-identifiable part of the combined trademark is removed and the distinguished part of the remaining trademark is called the main part. Each of these points can be subject to trademark similarity judgment and registration possibility judgment.

Language processing system

Concept and Necessity

Trademarks filed in the Republic of Korea often contain foreign languages as a component. This is similar to the fact that each country includes non-Korean elements as trademark elements. In this case, when the country to which the present invention is applied is the home country, the native language may correspond to the home country, and when the language system of the native language and the language system of another country become trademark components, the problem of foreign language Occurs. It is described herein under the assumption that the present invention is practiced in the Republic of Korea for convenience of description. Therefore, the word "Korean" or "Korean" is used to represent the native language, and the word "English" is used to represent all foreign languages other than the native language. In the case of Korea, the native language may be Korean, and the foreign language may be English, Japanese, French, Chinese, etc.

The necessity of the language processing system is as follows.

First, it is a matter of trademark similarity judgment.

If this process is extended, language processing is required for the national language constituting the trademark in various trademark DBs (exceptional case for determining similarity based on foreign language) or foreign language. In general, language processing is also required in various dictionaries and other elements used in various judgment processes related to trademarks in the present invention. In other words, the language can be judged similarly only when there is a language processing process that unites or relates the language.

Second, it will be a problem in determining the possibility of registration.

Thirdly, because computers are inherently incapable of thinking, they are only able to perform various processes only when the standard of processing is unified (when there is relevant information, it is uniform based on the information). do. Therefore, the language processing system becomes a problem in some or all of the processes to which the present invention is applied.

Characteristic

When a trademark component includes a foreign language, it has the following characteristics.

First, the appearance is different. That is, King, king, and wang are similar in concept but are completely different in appearance. Conversely, the appearance may be different but the concept is similar. And even if the appearance is different, such as "tiffany" and "tiffany", the title may be the same or similar. There may also be two or more language expressions in a trademark.

Second, the title can be generated by looking at the foreign language part of the emblem containing the foreign language, and the title can be written in Korean so that it has a certain degree of 1: 1 correspondence in the Korean language. One-to-one correspondence means that one language can be matched to a foreign language as spoken or read, as in the case of “CHANEL” to “CHANEL”. Although a certain degree of correspondence can be pronounced as "CHANEL" in "CHANEL", it can also be pronounced as "SANEL" or "CHANEL", so one or more languages can be corresponded to a foreign language. It is said that Koreans represent the same foreign language without a big difference.

Third, when a foreign language is a word included in a dictionary, a part of a word, or a combination of words, a phonetic symbol or a Hangul-Roman (foreign language) notation existing in the dictionary may be used.

English of Korean

The most important factors in determining the similarity of labels are appearance and title. At this time, if the language constituting the emblem is different, the problem of similarity of appearance is greatly reduced, and the problem of title and concept is raised. This section focuses on the issue of title. The idea of analogy is not addressed here because it can be solved by looking up a translation or a Korean-English (Korean-English) dictionary. The Korean language of English determines similarity based on Korean language. However, if you go through the reverse process of mapping English to Korean, you can change Korean to English. In this case, similar judgments are made based on English.

Titles in English may correspond to one or more languages. Therefore, when English and Korean are to be compared, titles are mainly the center of comparison, and there may be a voice-to-speech comparison and a corresponding character-to-text comparison for voice comparison. Normally, the comparison of voice is to compare the voice data of the English emblem and the Korean emblem under specific criteria to determine similarity. Character comparison is to compare the titles corresponding to English with the Korean characters. If the Korean character is converted into a comparison problem between the Korean language and the Korean language, it is a question how to translate the English language into Korean.

Use of International Speech Symbols

Phonetic symbols in the dictionary if the English element of the trademark is a word in a dictionary or a combination of words in the dictionary (to the extent that the part of the trademark is a word in the dictionary) Based on this, the Korean character can be converted accordingly. Depending on the letter system and phonetic system of each country, it may not be possible to completely correspond to each phonetic character or phoneme to one phonetic symbol. That is, when there is a method of reading the specific phonetic symbol under the language system of the country, it is possible to match the phonetic symbol or symbol group to Korean phonetic notation corresponding to the reading method. The DB used in the process of implementing this is the phonetic alphabet corresponding DB.

In most countries, there is a notation based on the country's linguistic system of English, which is defined by the country or its authority to establish public, academic or other standards. By using these notations, the Korean language can be mapped to correspond to each English itself or the phonetic symbols of the English language. In other words, the DB required in this case is largely a dictionary DB and a notation DB.

Using Hangul Romanization

Each country usually has a notational system in which the names, names and other words of its own nationality are written in two or more foreign languages. An example is the Hangul alphabetic notation, and the corresponding English Hangul notation exists in the United States. There may be several such notations, such as national standardized, de facto standard, made by an individual or a specific organization, and the present invention utilizes some or all of them.

Use nearest vocabulary phonetic symbols

If the English is not in the dictionary but a part of the English language constitutes a part of the word existing in the dictionary, the Korean characters can be mapped to the English language. When the part of the English becomes part of two or more words instead of a part of one word existing in the dictionary, two or more Korean characters may be formed based on phonetic symbols and the like corresponding to a part of each dictionary. The same applies to the case where two or more portions are present in English. In this process, if English is not fully processed and part is left, it can be converted into Korean language by using the Hangul alphabet notation, or by using phonics or other parts of the dictionary closest to the rest. Can be.

How to use segment symbols

Many English (used to represent English in general) are sometimes marked with phonetic units (syllables, etc.) in dictionaries and Korean dictionaries. In general, syllables indicate that the center point (in this specification, the center point in the notation uses “·”), spacing, or any other means of identification means that a spelling group of a certain unit becomes a pronunciation unit.

For example, if there is a word "identify", if there is a segment symbol "i den" ti "fy", the phonetic symbol corresponding to this segment symbol is often segmented. For example, the following notation may be used. [ai, den, ti, fai] In this case, the segmented spells I, den, ti, fy, etc. are referred to as the segmentation order group, and the phonetic symbols ai, den, fi, fai, etc. are referred to as segment pronunciation. .

Korean pronunciation may be displayed in response to segmental pronunciation. For example, [ai] can correspond to [child], [den] to [den], [ti] to [ti], and [fai] to [pie]. These [ai] etc. are called "fractional pronunciation notation Korean pronunciation notation".

Corresponding relationship between English notation and Korean notation can be obtained by following the following process with Korean phonetic notation.

First, a plurality of words are extracted together with a pronunciation from an English dictionary marked with segments. Sort by each segment spelling group while maintaining the correspondence between the segment spelling group and its segmentation notation. Through this process, you can find the various spelling groups that make up the English words, and find the corresponding phonetic marks for each group. In this case, the same segment spelling may have different pronunciations in the group, in which case a weight based on statistics may be applied. That is, if one segment pronunciation notation has two pronunciations, and each of them is in the ratio of 70% to 30% according to the analyzed data, the two segment pronunciation notations correspond to one segment spelling group, You can assign weights of 70% and 30% (of course, you can ignore 50 and 50%).

Second, the Korean pronunciation may be corresponded to the segmentation notation. The Korean alphabet may correspond to a phonetic symbol constituting the segmental phonetic notation, or the Korean syllable may correspond to the phonetic notation itself.

Third, through the above process, it is possible to correspond to each segment spelling group, segment pronunciation notation, and its Korean pronunciation notation. In this case, if the segment group has two or more pronunciations, the Korean phonetic notation may be corresponding to each of them, and the weight may follow. This final notation is called "segment spelling group corresponding Korean notation".

If there is an English notation as a component of a trademark, the application of this segmental spelling group corresponding Korean notation is as follows.

First, segment your English to include as many segmented spelling groups as possible. (Of course, you can use the English dictionary (including the human dictionary) to find the longest word in the English notation and make the first segment by word. It may be changed to Korean notation corresponding to the segmented segment spelling group. If it is possible to segment in a plurality of ways, it may come out in a plurality of corresponding Korean notation. Of course, it may also be output in the corresponding Korean display to which the weight (based on the probability) will be given.

How to use other correspondence

You can also use other pronunciation and character matching rules. Word smart, published by Princeton Review Publishers, displays the pronunciation of each English word in its own unique system of notation. As such, a new arbitrary notation system can be conceived for expressing sounds for each language, and those marked with these notation systems can be targeted for trademark similarity judgment.

Generalization of Correspondence and Application of the Invention Idea

By using the above methods, the title portion of the trademark in English can be mapped using the phonetic notation of Korean, Japanese, and French.

First, in the case of English, other English may correspond to the place, and the Korean language of the country to be translated or converted may correspond to the place of Korean. That is, in the case of carrying out the present invention in the United States, in the case of a trademark including a Japanese filed in the United States, English, Japanese, and Korean, English may be entered in Korean.

Second, the process of figuring out which language is used to apply the above invention idea can be combined. Korean, Japanese, Chinese, etc. are not a big problem because the notation of the language system is different, but if the alphabet is a notation system may cause problems. It can specify a country if there is a characteristic of each language's writing system (eg umlaut markings in German). If it is difficult to specify a country, the country in which the present invention is applied has the least frequent order of languages using the notational system (here, alphabet) (for example, English, French, German if the alphabet is used in Korea). To specify a country.

Where and how to solve language-related issues

At present, a number of trademarks in foreign languages or foreign character systems are used, which is a universal phenomenon worldwide. Problems related to language are as follows, and the solution is as follows. Unless otherwise specified herein, language processing may already be assumed.

Trademark DB establishment

When building a trademark DB, a trademark in a foreign language can be selectively constructed in Korean. In other words, it can look at the appearance of the trademark and can write the pronunciation of the trademark can be written in Korean. The Korean Intellectual Property Office stores a trademark component in English with the Hangul notation written as it is spoken.

Trademark Similarity Determination (5-1)

In determining similarity, a trademark in a foreign language may be translated into Korean and Hangul, or may be judged as a foreign language or a foreign language. If any foreign language trademark has been translated into Korean by an official such as the KIPO, it may be consulted first. When judging from Hangul versus Hangul, various Korean pronunciations of one foreign language trademark can be written, and thus, each of them can be determined. At this time, the pronunciation of each Hangul pronunciation for a foreign language can be timed and the similarity determination process can be performed under that notation.

Registration possibility judgment

When storing a common name, a conventional name, a geographical name, etc. in a DB, the main foreign language notation including English may be written in parallel, or the main foreign language notation viewed from the viewpoint of the country in which the present invention is used may be stored in parallel. If there is a trademark in a foreign language, it can be judged by the Korean language by the Korean language, or judged from foreign language to foreign language.

Unidentifiable component

There may be some non-identifying components in foreign languages. For example, Co. Ltd, etc., these may be stored as non-identifying components DB as non-identifying components.

Product similarity judgment system (200 / 1-1)

Introduction

In general, although similar marks are similar, certain legal effects may not occur in the legal system related to the mark such as trademark law or unfair competition prevention law, if the object using the mark is not similar. Therefore, in the present invention, the same and the like of designated goods are dealt with together. However, each country's Patent Office and Patent and Trademark Office establish standards for product classification, and NICE classification system is adopted in many countries internationally.

The principle of similarity between designated products is to be a judgment that considers each country's legal system, precedents and examination practices. However, in the present invention, judging by considering one or two or more of the application of the classification system of goods in the examination practice of each country, the NICE classification system, or the classification of merchandise.

Product Similarity Determination Engine 200-1

Concept and Necessity

The product similarity judgment engine 200-1 refers to the similarity of a product (including similarity of service industry) and the similarity of a product and a service industry (which is generally regarded as a problem of “homogeneity” in the Korean trademark industry). It consists of various programs and a database (DB) for judgment. The programs relate to theories and applications of artificial intelligence theory, linguistic engineering, mathematical modeling techniques, and other computational methodologies to determine similarities.

An external third party may access the product similarity determination system 200 / 1-1. At this time, the external and registration possibility determination system 400 / 1-1 may use the following information: 1) the product similarity determination system 200 / 1-1. This server consists of 2) at least one user (applicant, customer, member, etc.) terminal, 3) a network 1000 system connecting them.

The product similarity determination engine 200-1 is configured as a module for forming a similarity determination rule, a product DB 770, and an application module for applying the similarity determination rule, and interact with the product similarity determination rule DB.

How to Form Product Similarity Determination Rule 200-2

In order for the computer to automatically determine the similarity of the goods, the product similarity judgment rule 200-2 is required. There are three ways to make the product similarity determination rule 200-2.

First, it is a method that utilizes various classification systems set in advance for the convenience of legislation or administration such as designated products, designated services, or NICE by trademark-related administrative agencies such as the Patent Office.

Second, the merchandise classification system, the HS code classification system, or other product classification systems.

Third, judicial organizations such as the Supreme Court and administrative agencies in charge of trademarks are used.

Fourthly, one or two or more of the above three methods are integrated.

Stage of similarity judgment rule 200-2

Type of Rule Group

First, a random setting step of the product similarity judgment rule 200-2. This rule group consists of three major groups.

1) As a general rule, for example, under the Product Classification Criteria Table under the Korean Trademark Act, “A different kind is a dissimilar product. Products belonging to the same group of the same class are similar products. ”

2) As a concrete standard, it is a method to determine the similarity between specific designated products by using specific examples determined by the judiciary agency such as the Supreme Court and the administrative agency in charge of trademark affairs. For example, if the Supreme Court judges that a futon and an electric futon are dissimilar, then certain designated commodities, futon and electric futon, may be dissimilar.

3) A new standard can be set by combining the above. For example, if a futon and an electric futon are dissimilar, other goods in a certain group of the particular class to which the futon belongs to and other commodities in a specific group of the particular class to which the electric blanket belongs are set similarly.

Expansion of Rule Group

Expanding the above logic is as follows.

In general, it is a method of identifying similar products belonging to a lower class below a certain level in product classification, NICE classification, or any other classification system, and identifying similar products belonging to a lower level below a certain level. This classification criterion may have a domestic or foreign classification.

A job is usually closely related to the services it provides. In the United States, there are about 300,000 occupations classified and classified, and the Republic of Korea is classified into about 50,000. In this case, a corresponding relationship may be made between the classification of the occupation and the classification of the designated service industry, and in this case, the occupation classification may be used to determine the similarity of the designated service industry. In other words, if the similarity of a specific designated service industry is a problem, this fact may be the basis for determining the similarity of the designated service industry when the occupation or occupational group corresponding to the designated service industry has the same, similar or common parts. .

2) Specific standards should respect general standards, but can supplement specific inconsistencies that occur when applying general standards by using specific examples determined by law enforcement agencies such as the Supreme Court and administrative agencies in charge of trademark-related work. . In other words, the application of the standard is to apply the general standard first, and not to conclude with the general standard alone, to examine whether there is a specific standard, and to output the conclusion according to the general standard if there is no specific standard.

That is, in the step of setting the similarity judgment rule 200-2 group, it is necessary to set the similarity judgment rule 200-2 group, and to find an exceptional specific example that is inconsistent with this rule, and to store, manage, and apply them as a DB. It is the key.

Distance weights

The trademark law stipulates that classification of products is not a basis for judging similarity, but the classification table reflects knowledge of product similarity based on decades of examination practice. Therefore, it is possible to actively use this product classification / classification table DB 720-1 when automatically identifying product similarity, one of which is the distance weighting method. The location of a particular product on the classification table is 1) within the classification unit, 2) within the group classification unit, and 3) within the group. In addition, products with high similarity are often located close to the military classification. In addition, the number of designated products in one group may be large or may not be many. In many cases, there is a commodity segmentation, and this proves that the market is bigger or bigger. Therefore, when judging the similarity of a particular product, one may choose to examine the distance between the two products. In other words, if the distance between the two products in the group is far apart, the relative dissimilarity may be high, and may also be applied between the group within the group.

Distance between two products on the merchandise taxonomy / categorization DB 720-7 = sum {(distance between the highest classifiers * weight for the highest classifier) + (distance between the lowest classifiers * weight for the lowest classifier + (Distance within top class * weights within bottom class)}

Each weight may vary and may be modified to reflect the specific judgment result derived from the examination, judgment, litigation, and the like.

Simultaneous Use of Classification and NICE Classification

Using distance can also be used in the following ways. If there is a designated product to be prepared, this product can be judged at the same time by applying similar judgment criteria in the new classification and classification. If the distance between the products is close on the classification and the distance between the products on the new classification, it can be judged that the degree of similarity of the products is statistically higher than that of the products. That is, whether or not the product similarity criterion is applied (for example, whether it is in the same group) or the distance on the new classification and the classification, the total of the distance between the two classifications or a certain weight is given to each classification criterion (new The smaller the sum of distances (60% for classification and 40% for classification), the higher the similarity of goods.

This concept can be extended to other product classification criteria. For example, you can check the distance or degree of similarity between two products on the basis of HS code, and calculate the total of distances between classifications, new classifications, etc., or the total distances given a certain weight to each classification criteria. The degree of similarity of the product can be determined. Designated services can be matched with a occupational classification system, such as a occupational classification dictionary or occupational yearbook.

If there is a classification of the product classification table / classification table DB (720-7) and the NICE classification, there may be a connection or correspondence of a specific product. Using this, information similarly judged by various precedents and judgments on the classification can be applied to the NICE classification.

Weight setting step

Second, in this case, when there is data that the products of these two groups are judged to be similar in another case, that is, when there is a case against the judgment of similar or dissimilarity in the military group, the similarity of the group is determined. For example, a method of setting weights (0 to 1, weights may be set by reflecting the number of conflicting cases or judgments) may be used.

This will be described in detail as follows. If there is data judged on the similarity of goods, 1) check the group to which each product is compared, and check whether there is any data judged by precedent etc. for each product belonging to both groups, For example, if there are 10 designated commodities in the group to which the futon belongs, and if there are 20 designated commodities in the group to which the futon belongs, then a combination of 10 * 20 commodity groups will be produced. 2) If there is judgment data, the number of positive and negative similarities can be identified and weighted based on the similarity (for example, 5 out of 200 pairs). If there is a comparison of pairs of dogs, one of the five affirms similarity and four deny the similarity. A method of giving such values.)

Expanding the above logic is as follows.

According to the criteria of the Korean trademark examination, if the classification is different, they are not similar, and if they are the same within the same classification, they are considered similar. However, the case is judged by the general consumer's attention to reflect the specific status of the trading society to determine the similarity by whether misunderstanding of the source occurs.

In this case: 1) If there is a relationship between different subclasses that exist in the horizontal level under the same higher class (for example, classification of goods) under the same classification system (for example, classes are the same and different groups), 2) (3) In the case of a relationship between a (designated) service industry that is highly related to a (designated) product, and 3) the above logic may also apply to the case of both goods under different classification schemes, and in the case of goods and services.

Establish weights for these rules by setting general rules for similarity between specific comparison targets (e.g., product A and service sector), and analyzing and verifying how well they apply. The steps can be performed.

DB required

The following DB is required for product similarity judgment (5-2).

First, to the product and service industry, such as the DB (including classification and new classification DB) for the designated product and the designated service industry, the product classification DB on the HS code / code DB 720-3, the merchandise product classification DB, and the occupation classification DB. DB is required.

Second, it is a DB that reflects information on a particular product or product group (service industry, service industry, and product and service industry) that are recognized or denied similarity in the examination, judgment, litigation, etc. In this case, weights may be differently applied to each judgment such as an examination, a judgment, and a lawsuit at each stage. In other words, the weight of the judgment shown in the Supreme Court case may be different.

Third, it is a DB related to the product similarity judgment rule 200-2. This DB may contain information about the product similarity determination rule 200-2 and specific practical examples contrary to the rule. These rules may be weighted for the formulas that apply them and the application of those rules.

The first DB is the narrowest product DB, and if the first and second are included, the broadest product DB is the broadest product DB. In the description of the present specification, the product DB uses the term product DB as the broadest concept for the convenience of description without special mention.

Product Similarity Determination (5-2) Processor

The product similarity judgment (5-2) method is as follows.

Basic course

First is the specific process of goods. In the case of a normal trademark application, the designation process of the designated product corresponds to this, but it is possible to specify the product without presuming the application. The methods for specifying the goods can be largely as follows.

It provides a product classification table / classification table DB (720-7) to the user of this system to specify the product while going through the order, such as class-gun-product details.

2) Enter a keyword or a product desired by the user, and display the singular or plural products or the product or group of products to which the product is optimally matched and input the final selection of the user. How to receive.

3) Provide a product classification table / classification table DB (720-7) different from the product classification table / classification table DB (720-7) (for example, HS code / code DB (720-3) classification table, etc.) The product may be specified by matching the product classification table / classification table DB 720-7 with the product classification table.

Second, the specified product is referred to the product similarity determination engine 200-1. This inquiry process may occur during the specific process of an individual product, but may also occur after all products have been specified. The product similarity determination engine 200-1 queries the DB regarding the product similarity determination rule 200-2, and searches for a recognized product or related service business that is identical or similar to the specified product. This process can be done together with or independently of the trademark similarity judgment 5-1. That is, a trademark similarity judgment (5-1) is made before or at a specific moment of the product first, and a product similarity judgment (5-2) is performed within a range of the designated product of the trademark for a trademark whose similarity is recognized. You may. In addition, although there may be a similarity determination of the trademark prior to the similarity determination of the merchandise, the similarity determination of the merchandise may be performed first.

Third, the product DB is queried based on the product similarity determination rule 200-2 to determine the similarity of the specified product.

Extension process

Various recommendations may occur in the specific process of the product. When the user selects one or more products, the recommendation may recommend similar products, related products, or designated service businesses to the user. The criteria for recommendation are: 1) various designated goods or designated services within similar ranges of the product selected by the user, and 2) designated goods or designated services, reflecting information on which other products or designated services are selected by those who select the product. .

Formation and Renewal Process of Rules for Determining Similarity between Trademark and Product

The process of forming and updating the rule of similarity between the trademark and the product will be described with reference to FIG. 4.

First, a trademark similarity judgment rule is extracted (4-2) based on a similar trademark collection / similar trademark DB (710-4) or various DBs related to similar trademarks (710-4a) or a trademark similarity judgment rule is input (4). The weight of the trademark similarity judgment rule is set in the manner of processing by -1). (4-3) This forms the trademark similarity judgment rule DB. (4-4)

Second, the product similarity judgment rule is extracted based on the similar product collection / similar product DB 720-5 or the various DBs related to the similar product 720-4a (4-6), or the product similarity judgment rule is input (4). The weight of the product similarity judgment rule is set in such a manner as to process the product similarity rule. (4-7) This forms a product similarity judgment rule DB. (4-8)

Third, enter this rule into the update engine. At this time, the update engine enters the examination / decision / preceding data (4-10), the JPO announcement data (4-11), the rejected data (4-12) or / and other data (4-13). Based on these data, the update engine generates an updated trademark similarity judgment rule DB 760 (including an optionally updated weight) and an updated product similarity judgment rule DB 770, and includes a trademark DB, a product DB, and the like. Update other related DBs.

Similarity Determination Processor

The similarity determination processor is as follows.

Trademark / Product Entry Steps (2-1)

Enter your trademark and merchandise (2-2). There are two ways to input.

The first is that human input. A person may enter a title or the like of the trademark in letters in a trademark in which figures and the like are combined. In this case, there may be several ways of inputting one trademark, and each of them is determined to be a valid input.

Second, it reads from DB and inputs automatically. In this case, if the information on the character part is added (the DB provided by the Korean Intellectual Property Office has information on the character part in connection with the image data of the trademark), the character part is automatically accepted and entered. The letter portion can be recognized from the trademark image data through the method of OCR (Optical Character Recognition).

Processing step (2-3)

Similarity judgment is largely a product similarity judgment (5-2) and a trademark similarity judgment (5-1), so they can be processed at the same time, or a method of judging either one first. Both differ only in the order of internal processing, and the results observed externally may be identical. Usually, designated products are often designated from the list of designated products, and the batch processing regulations (such as non-like products if the classification in the JPO's examination criteria are different, and presumed to be similar products if the groups are the same) Since there may be cases, making the product similarity judgment 5-2 first can reduce the amount of data processing. On the contrary, there is a method of searching for similarity of trademarks first and searching for similarity of goods. Powder it below.

Unilateral and bilateral

It is called one-sided problem to judge similarity and possibility of registration in comparison with a trademark of an unspecified majority (mainly DB) about one trademark. This is mainly a matter of application, examination, objection or invalidity judgment. In this regard, both problems refer to the comparison of one trademark with one or more specified trademarks. Hereinafter, it describes based on one problem.

Product priority processing (product processing process (2-3))

The first step is to read the target trademark and the specified designated product. When there are a plurality of designated products, the above and subsequent steps are repeated for each product. The designated product is compared with the product classification / division table DB 720-1. If the product is in the product classification / classification table DB 720-1, check the identity with the product in the product classification / classification table DB 720-1. It goes through a "statute-designated commodification process" to find the nearest product.

Second, a product that is recognized similarity with a specific designated product or a designated product that has undergone a legally designated commercialization process is determined in the product classification / division table DB 720-1.

Third, the product DB is searched for products that are recognized for similarity. Here, the product DB is due to the Office of Patent and Trademarks, which are trademarks that have been disclosed or registered (or which have not expired after one year of registration) (due to Article 7, paragraph 1, 8 of the Korean Trademark Act) and their trademarks. It refers to a database in which a designated product is stored in relation to it, whether the DB is inside or outside the KIPO. Extraction of Trademarks This process creates “existing pairs of trademarks”, each of which is a component of a product and a trademark for which similarity is recognized.

Fourth, the similarity between the target trademark and the existing trademark is determined through the trademark similarity determination system 100 / 1-1. Since the similarity of the goods has already been determined, only the similarities of the trademarks matter.

Fifth, collect the results. The target trademark is compared with all existing trademark pairs, and the comparison result is stored in the DB according to the processing method of the similarity judgment system 300 / 1-1. If only one of the existing trademark pairs is similar, the result should be similar, and if the judgment of similarity is above a certain level (if the internal standard is 70%, the determined degree of similarity is above 70%). For all data that exceed, you can print from the one with the highest similarity.

Brand Priority Process

The first step is to read the target trademark and the specified designated product.

Second, trademarks that are recognized for similarity with the target trademark are extracted from the trademark DB. At this time, it can be arranged from a trademark having a high degree of similarity.

Third, the designated products of the trademarks are extracted for the trademarks that are recognized for similarity. For trademarks that are recognized for similarity, "existing pairs of trademarks" are created, each of which is a component of the product and trademark.

Fourth, the similarity judgment between the designated product of the target trademark and the product is determined through the product similarity determination system 200 / 1-1. Since the similarity of the trademark has already been determined, only the similarity of the product matters.

Fifth, collect the results. The designated product of the target trademark is compared with all the designated products of the existing trademark, and the comparison result is stored in the DB according to the processing method of the similarity judgment system 300 / 1-1. If only one of the existing trademark pairs is similar, the result should be similar, and if the judgment of similarity is above a certain level (if the internal standard is 70%, the determined degree of similarity is above 70%). For all data that exceed, you can print from the one with the highest similarity.

Miscellaneous issue

Applicant's identity

If the applicants are the same, the determination of similarity is unnecessary unless it is a matter of trademark 1 application. Therefore, if the information on the applicant or the owner is not a problem from the beginning in the similarity search of the trademark, the process is described separately.If the applicant or the owner is specified, the above processing is performed except when the applicant is the same in the similarity judgment data do.

When there are a lot of designated products

When there are a number of designated products, they are logically essentially the same as when there is a designated product, but under some specific circumstances there are ways to speed up the processing. To this end, when there are a plurality of designated products, it is possible to determine where the designated products are arranged in the product classification / division table DB 720-1 and the attributes (such as the same group of the same class) of the arrangement.

First, it is the processing of the smallest group unit where the similarity of the designated product, which is the basis of internal examination by the KIPO, is recognized. That is, when the designated product belongs to the same minimum group unit, the minimum group unit can be treated as one product and processed at once.

Second, if the minimum group units of the designated products are different, the first process described above is performed for each of the minimum group units.

Statutory Commercialization Process

The court-designated merchandising process refers to a merchandise classification / delimitation table when an applicant or the like inputs his / her designated product arbitrarily, rather than specifying a product in a legal merchandise classification / delimitation table DB 720-1. It refers to a process of changing to a statutory designated product in the DB 720-1. This process is related to Article 10, paragraph 1 of the Korean Trademark Act and applies even if the foreign trademark law has the same purpose. This process may occur during the filing process, or the applicant may search for a trademark on the network 1000. In particular, instead of specifying the designated product by selecting a product in the product classification / division table DB 720-1 (selection method), the applicant automatically inputs the designated product when the applicant inputs the designated product. More useful in the extraction process.

First, it is checked whether the designated product inputted by the applicant or the like is a product on the product classification / division table DB 720-1.

Second, if the product is not on the merchandise classification / classification table DB 720-1, the merchandise entered by the applicant, etc. and the nearest merchandise on the merchandise classification / classification table DB 720-1 are searched for. The method of finding is to find the place where the inputted product is located in the product DB or the product science dictionary, and the same, similar or common part is identified with the product category of the upper category of the place where the product is located or other products in the equivalent category. The product which is higher than the standard is searched in the product classification / classification table DB 720-1. 2) If the product entered by the applicant is a higher concept of the designated product on the merchandise classification / classification table DB 720-1 (for example, "shoes" is not present on the merchandise classification / classification table DB 720-1). Instead, there are specific sub-category products such as soccer shoes, tennis shoes, etc.) Finding a category with a higher classification of the product entered by the applicant in the product DB or the product science dictionary, and classifying / dividing the products within the category with the products Contrast with the goods on the table DB 720-1, or present the product classification / division table DB 720-1 to the applicant or the like.

Third, the applicant selects and confirms the designated product.

Result output step (300-4)

Representation of Similarity Judgment

Since the similarity between trademarks is closely related to the similarity between products, if there is a trademark and a designated product, the similarity determination process proceeds as follows. In order to be confused with both trademarks in question, the trademarks must be similar and similar. Therefore, one, two, three, or N-dimensional matrices expressing the data structure of the judgment on the similarity judgment or the manner in which the data structure is revealed or the processing of the similarity judgment to create the data structure and the result of the processing. ) Can be thought of.

1-dimensional matrix

The one-dimensional matrix refers to determining only the similarity of only one of the trademark or the product. In other words, if there is a trademark or product in question, the trademark or product that is most similar to the trademark or product may be arranged in one column on a logical level. The elements that make up each array may contain evidence of similarity.

Evidence data may include essential similarity judgment rules, whether or not exceptions are applied to the rules, precedents related to the exceptions, and various additional data that can increase the reliability of various statistical values and similarities. In addition, when a user requests a printout of the evidence, the evidence may be separately output to the user in a report form.

Two-dimensional matrix

The two-dimensional matrix refers to determining overall similarity that simultaneously identifies similarities between trademark and product. At the top left, there may be the trademarks most similar to the particular trademark in question, and the designated products most similar to the designated goods, and vice versa, arranged at the bottom right of the right. Small arrays can result in arrays.) The elements that make up each array can be combined with supporting evidence for similarity judgments.

3D Matrix

Three-dimensional matrices can be used when there is a question of the tightness or relationship of goods and services. That is, when a specific designated product has a close relationship with a specific designated service industry, including a similarity problem with another product, the similarity with the designated service industry needs to be considered together. Therefore, it is necessary to grasp the three-dimensional similarity with the concept axis in the height direction.

n-dimensional matrix

In case of using different product classification criteria (service industry classification criteria), applying only a specific part of the trademark or product similarity judgment rule 200-2, or determining the similarity by changing the weight of each rule in some or all of the rules. The matrix requires three or more dimensions. In this case, however, the logical processing method is the same as above.

Comprehensive judgment expression

In the above dimensional matrix representations, the results of the comprehensive judgment of the similarity between the trademark and the product can be collectively expressed in the order of similarity. There may be numerical (80% similar, etc.) or gradual expressions (identical, very similar, highly similar, etc.) of similarity in this comprehensive expression. Can print In addition, the information on the counterpart trademark to be compared also includes any expression means for displaying the counterpart trademark (the trademark image itself, a unique number (application number, serial number, registration number, etc.) for the trademark, and other trademarks thereof. : 1 In the case of acting as a click on any corresponding means), any information on the trademark can be extracted from the DB and output.

Report Writing System (300-5)

The reporting system of the similarity judgment result is any system related to how the output of the similarity judgment result is expressed. If there is only one designated product, it can be expressed as above. If two or more designated products are selected, the result can be output as follows.

First, as a result of comprehensive judgment on the similarity, the similarity can be output in descending order.

Second, the result of the basis of the similarity determination can be output as the user's request or default. The similarity determination engine may present the user with information on how well the similarity determination rule is applied to the trademarks and products to be input during the application of the similarity determination rule.

Third, if there is an existing trademark that the similarity is recognized to a certain degree or more, information about this can be presented to the user. To this end, 1) an arbitrary number of unique identifiers such as an application number and registration number can be used to inquire in real time to the DB of the KIPO, or 2) if information on an existing trademark is found in a trademark DB, the information can be extracted from this. Can provide. In the case of 2), the contents of the trademark DB may be different from that of the JPO, so it may provide information to the user after checking with the JPO DB.

Trademark Registration Possibilities Determination System (400 / 1-1)

Concept and Necessity

A trademark registration possibility determination system (400 / 1-1) that provides a system for automatically and semi-automatically examining whether or not a registration requirement is satisfied for a particular trademark registration application by providing a judgment on the possibility of registration and the basis for the determination. This is called.

In the Korean Patent Office, the trademark registration rate is only about 30% compared to the trademark registration application (1999 statistics), and about 70% of the trademark registration applications are filed with a considerable number of applications that do not meet the registration requirements.

The judgment of trademark registration requirements is a high level legal judgment and cannot be ruled out by the judge. However, the present invention can not only provide evidence or refusal evidence for their judgment, but also help the convenience of searching at the level of users who lack expertise in general law, Make a first or second judgment about this, saving you time and money.

Trademark registration requirements vary from country to country. However, the present invention is described based on the trademark law of Korea and Japan, and even if each country has a different legislation, essentially including similar judgments of trademarks and designated products constituting a part of the registration requirements. That is, each country's trademark law includes some or all of the trademark registration requirements described below. Under the laws of each country, some or all of the following may be used to determine the possibility of trademark registration.

Trademark Registration Probability Determination Processor (400 / 1-1)

The similarity determination / registration determination process will be described below with reference to FIGS. 5 and 6.

First, there is a process of inputting a trademark / good (2-2) and processing of a trademark / good (2-3). In this regard, a trademark similarity judgment (5-1), a product similarity judgment (5-2), and a result of the judgment are prepared in a report (5-3).

Second, independent of the first process, it is possible to carry out the registration possibility determination process (5-4) and to prepare a registration possibility determination report (5-5).

Third, when the report is output (5-6), the report is transmitted and the application intention from the recipient (5-7) is received, the online electronic filing process (5-8) / (530) can proceed. have.

6 is a conceptual flowchart for a registerability determination processor.

First, there is input of a trademark / designated product (2-2), and processing for the trademark / product (2-3).

Second, a trademark / product is entered into the detailed registration possibility determination engine 11-1 for the first registration requirement, and the determination engine determines the detailed registration possibility determination rule DB (11-3) and the first registration for the first registration requirement. Using the detailed registration possibility judgment data DB (11-2) on the requirements, it is determined whether registration is possible by the first registration requirement (11-4), and a judgment basis report is prepared (11-5).

Second, perform the first decision process (11-6) above for final registration requirements from the second.

Third, the results of the detailed registration possibility determination engines for each registration requirement are synthesized (11-7), the comprehensive judgment on whether registration is possible (11-8), the judgment basis report preparation (11-9), and the results are output ( 11-10).

This will be described below.

Configuration

A description with reference to FIG. 7 is as follows.

The registerability determination system 400 / 1-1 is largely divided into a registerability determination engine 400-1, a registerability determination DB 400-2 (detail registration possibility determination rule DB 400-2), and a detailed registration possibility determination. Data DB (400-2a ~ 400-2n)) and reporting system. Since there are a plurality of trademark registration requirements (trademark law should not be the reason for rejection in a broad sense, and in consultation the requirements of Articles 6, 7, and 8 of the Trademark Law), there are also a number of detailed engines that judge the requirements. There should be dogs.

This system should, of course, include the trademark / goods entry process (2-1), trademark / goods entry (2-2), trademark / goods processing (2-2), trademark DB (700), judgment There is a result output system 300-4, a report preparation system 300-5, and an online electronic filing system 300-6 may combine.

An external third party may access the registration possibility determination system 400 / 1-1, whereby the external and registration possibility determination system 400 / 1-1 may be: 1) the registration possibility determination system 400 / 1-1. This server consists of 2) at least one user (applicant, customer, member, etc.) terminal, 3) a network 1000 system connecting them.

Registerability Determination Engine 400-1

The engine that determines the registration requirements of the trademarks in each country's trademark law is called a registration possibility determination engine 400-1. This engine combines and processes the results from the detailed registration possibility determination engine 400-1a to 400-1n and each of the detailed registration possibility determination engines 400-1a to 400-1n to determine each detailed registration requirement. It consists of.

DB with filtering

Filtering refers to extracting and removing a certain trademark by satisfying or not satisfying a desired condition by searching a DB having a certain property. And, to perform filtering, data DB is essential. The trademark and the designated product are inputted into an engine that determines each registration requirement (actually, a non-registration requirement) and proceeds in a manner of determining whether or not it corresponds to the requirement.

For example, suppose a trademark registration requirement is that a trademark should not consist solely of distinguished geographical names. In order to know whether a given trademark falls under this requirement, a database with a prominent geographical name is required, and an engine for determining whether the trademark corresponds to a prominent geographical name is needed. In addition, the reporting system outputs the result of this judgment.

Working processor

The operation method of the registration possibility determination engine 400-1 is as follows.

First, it is a step of receiving a trademark and a designated product. At this time, the processing of the trademark may be performed (2-3) before inputting.

Second, the registration possibility is determined by each detailed registration possibility determination engine 400-1a to 400-1n.

The names of the detailed registration possibility determination engines 400-1a to 400-1n may be referred to as "engineership mark determination application engines." An example of a rule that constitutes a rule DB may be "a common mark when the trademark (ruby) is a sub-component of the classification to which the designated product (apple) belongs in the product DB." Detailed registration possibility judgment data DB400-2a-400-2n becomes a product DB. In this case, the application result of the specific detailed registration possibility determination engines 400-1a to 400-1n may be digitized.

Third, the result of each detailed registration possibility determination engine 400-1a to 400-1n is synthesized to determine whether registration is possible. If the possibility of registration is applicable to any one of each detailed registration possibility criterion, the possibility of registration is denied. Therefore, the result can be weighted accordingly.

Fourth, the determination result is output. In addition to the comprehensive judgment at the time of printing, a report on the judgment of each detailed requirement and the basis of the judgment may be provided to the applicant.

Special case

If a preceding trademark is a concern, a judgment can be made regarding the possibility of conditional registration.

For example: 1) A pre-registered trademark may provide information that the pre-registered trademark is pending an invalid trial or cancellation trial, in which case the consequences of application of the law may vary depending on the outcome of the trial or litigation. The report can be synthesized and output to the applicant.

2) It may be less than one year since it is expired, but a report on how the law is expected to expire within a certain period of time or other extinguished trademarks works in relation to the trademarks of the applicant or the like may be printed.

3) If an application trademark is a problem, the application that has expired under the trademark law has no effect of excluding sponsorship. Therefore, the applicant can compose and output a report to the applicant, etc. by scenario on the legal status, effect, or possibility of future occurrence of the application. Can be.

Registration possibility judgment DB (400-2)

Basic registration possibility judgment DB (400-2)

The registration possibility determination DB 400-2 may include the following elements.

First, detailed registration possibility determination data DB (400-2a to 400-2n) for each detailed registration possibility determination engine (400-1a to 400-1n).

Second, each detailed registration possibility determination rule DB 400-2. This decision rule is composed of rules for determining whether or not to apply to each detailed registration requirement, such as trademark similarity decision rule 100-2. For example, in the case of prominent geographic names, an example of the rule is 1) rules for judging whether or not they correspond to items on the judging criteria of each country, 2) rules for judging whether they correspond to abbreviations of geographic names, and 3) In the case of a prominent geographic name in a foreign country, there may be many Koreans expressing it. Therefore, there may be a rule for determining whether the standard name is within a certain level of similarity (eg, Paris-Paris, Paris).

expansion

learning

Learning mechanisms that are important in artificial intelligence theory can be used to update each database and to refine the judgment of registerability. If there is a judgment on whether to register in an application review, trial, or lawsuit, the materials can be learned for each registration requirement. For this purpose, there is a learning routine utilizing artificial intelligence, and based on the input data, the DB update is made and the registration mechanism determination mechanism is further refined. As this process accumulates, the mechanism for determining registration requirements becomes even more substantial.

Reporting system

The configuration of the reporting system is conceptually the same as the similarity determination system 300 / 1-1. However, since there are a plurality of registration possibility determination elements, they may be divided into a system for reporting the determination result for each requirement for each requirement and a system for reporting the combined results.

How to judge by trademark registration requirement

Trademark registration requirements can be identified separately from general registration requirements and specific registration requirements. The general registration requirement usually determines registration requirements by considering only the relationship between the specific target trademark itself or the designated product of that trademark, apart from its relationship with other trademarks. The specific registration requirements are mainly the requirements for determining whether to register by making a comparative comparison under the relationship with other trademarks (you may also consider the relationship with the designated product of that trademark).

Common and Common Names

Common name

Whether it is a common name is judged as follows.

Trademarks entered into the registration possibility determination engine 400-1 (including processed trademarks and cut-off trademarks. If part or all of same) is same as designated product name, it is output as common name. Even if it is the same as that translated into some foreign language, such as English, it can be judged by a normal name, and even if it is the same as the upper category of the designated brand name. The common name is the common name if the trademark is identical to other language expressions of the designated product by querying the Korean or English-Korean dictionary for a given trademark.

Common emblem

To determine the existence of a common mark, a DB for a common mark is constructed through the following process. First is the destruction of the brand name. Trademarks are assigned by shredding them based on groupings of registered or pending trademarks and other trade names in use, grouped by each designated product, by product used, by product group to which the product belongs, and by product group to which the product belongs. The brand name is broken into syllables, words, words, and other designated units, and the broken syllables, words, words, and other designated unit names are assigned to each category by each designated product, used product, or product group to which the designated product belongs. When assigned, syllable groups correspond to each category. At this time, syllables, words, words, and names of other designated units that are overwhelmingly included in each category are likely to be common marks. These constitute a common emblem DB.

NICE classification / product dictionary / dictionary DB (720-2) / HS code / code DB (720-3) and common name common name

Products correspond to the NICE classification and the HS code / code DB 720-3, which are standardized worldwide. Therefore, by using the NICE classification or HS code expressed in each language, the corresponding foreign language expression of the product name can be obtained. In this case, you can find the common name in a foreign language.

If you have a more detailed product classification table than the NICE classification or the old product classification table, you can find common and common names. Examples of more detailed classifications may include an HS code / code DB 720-3 collection, a product information sheet (price information, etc.), a product dictionary / dictionary DB 720-2, and a domestic or foreign product classification collection. In addition, the classification codes used internationally are Harmonized System (HS), Standard International Trade Classification (SITC), United Nations Common Coding System (UNCCS), UN / SPSC (UN / Standard Products and Services Classification), NICE (International Classification of Goods Services), SKTC (Standard Korean Trade Classification) and EAN / UCC (European Article Number / Uniform Code Council) and Universal Product Code (UPC). These can be combined to achieve the goal. In addition, if there is a product to be compared on these classification systems, whether or not the similarity of the product can be confirmed through a method of evaluating the distance between the products.

In addition, it is possible to use a product studies dictionary or product information DB that classifies products in much more detail than NICE classification. Examples of the product information DB include the “price sheet”, the “occupation table” which shows the prices of various commodities, the product details for the survey of various consumer prices used by the National Statistical Office, and the other product classification table / categorization DB DB (720). -7) may be an example.

Technical emblem

Whether or not it is a descriptive mark is filtered through the following DB.

First, whether it is an absolute property label or not is used in various dictionary DBs for property labels. When transliterations of adjectives and verbs in the various languages of the language are used as transliterations (such as “resolve” in the detergent), it is possible that the appearance of a term or a changed form of a verb (these are called “convenient forms”) is a mark of nature. high. The following are necessary to determine technical marks:

Technical marks are often used as adjectives, verbs, or variations of them. In Indo-European languages such as English, adjectives, verbs themselves, or their common variations (passive, superlative, etc.) are overwhelming, and language systems such as Korean undergo severe modifications, unlike English systems such as English. Therefore, in order to judge verbs and adjectives, there must be an engine for judging which language is used and for processing a language adapted to the characteristics of the language. This is especially true of Korean.

First, it is a database of adjectives and verbs in Korean and foreign languages. This DB may include adjectives, verbs, and their conjugations in which all or frequently used Korean dictionaries are used. 2) In the case of a foreign language dictionary, different standards may be applied depending on the degree of recognition of the foreign language in Korea. In the case of a familiar language such as English, the elementary, middle school, high school or university level vocabulary (the elementary and middle school high school vocabulary is mandatory. (Of course, it can be applied to everything.)

Second is the terminology DB for each field. The specialized field may build a terminology dictionary based on a product group unit or a product group unit below it. That is, there is a classification system for each field in the technical terminology dictionary, etc., and the system may be used by matching the system with the merchandise classification / classification table DB 720-1. In this case, just by analyzing and statistically processing the telephone directory of each other and extracting the part without identification, each part of the field can be processed in a similar manner to find the part without identification only for the designated product related to the area of specialization. In the pharmaceutical field, for example, “pharma” is an example.

Third, thesaurus dictionary DB may be used to determine whether the property mark is indicated by the designated product. A thesaurus is a dictionary that combines synonyms, synonyms, related words, and related expressions for a heading.

Fourth, it is possible to use the trademark DB 710-3 rejected because of the technical marks. Since technical marks often express the nature or shape of a particular product, a technical mark is a technical mark if a particular product has its properties or characteristics, and if it does not have the properties or characteristics, it is likely to become a trademark that may mislead the quality or the consumer. Because it is high. This DB may be an example of domestic or foreign audit standards, reference materials for examination, data judged by a judge, a judge or a court.

Fifth, a DB can be constructed for each component of the technical mark of each country's trademark law. The place of origin is closely related to the geographical name, and the quality or efficacy is usually expressed as an adjective or verb itself or as a form of utilization. Raw materials can be obtained from a material related DB in commodity science. Since the vocabulary expressing the timing is limited, it can be processed with time-related DB. The production method, processing method, and usage method can be filtered by using the terminology dictionary DB in the case of verbs, verbs used or specialized technical fields.

A prominent geographical name

For example, one of the trademark registration requirements is that the trademark registration requirement does not consist solely of a prominent geographical name, map, or abbreviation. First, there must be a prominent geographic name DB that falls within the range of geographic names corresponding to the prominent geographic names defined in the law or examination guidelines. This prominent geographic name DB should include the abbreviation in parallel if there is a prominent abbreviation of that geographic name. (As for maps, etc., it is a matter of figure or figure, and not of text. This is dealt with in figure.) Overall information about geographical name can be extracted from map DB, postal directory DB, administrative area system DB or tourist destination DB. In addition, filtering this database by criteria found in legal or judging guidelines can create a significant geographical name database. Of course, human intervention may be necessary.

Common surname

Common surnames, names, etc. can be used by surname dictionaries from around the world. When this requirement is applied at each country level, it is determined by using the surname dictionary or telephone directory database of each country. If a foreign surname is also included in the judgment, a database of information on how many surnames are in the country can be used. This information DB can be determined by using a DB formed based on information on foreigners' surnames such as foreigners registered in government offices, visa issuance records, and tourist lists.

Other Registration Requirements

Violation of public law

Whether it is against public order or morals, you can use the review practice or slang dictionary database.

National flag

For the provisions of Article 7 (1) 1, 2, 3 of the Korean Trademark Act and the same purpose of foreign countries, detailed registration possibility determination engines 400-1a to 400-1n and / or its DB for each regulation Can be built. For example, prominent international organizations are often exemplified in the examination standards of each country's patent offices.In addition, DB for international organizations registered in the United Nations, etc., for only a certain number of registered members or a certain number of members is established. Can be. The same is true for public interest organizations.

Distinguished Brand DB

DB for domestic or foreign famous trademarks (hereinafter referred to as famous trademark DB) can be established. Prominent trademarks can be made through DB published by the patent office or customs office of each country.

Etc

In addition, a DB suitable for each registration requirement can be formed, and these various DBs can be created by inputting actual cases, judgments, judgment cases historically accumulated during the litigation process, and guidelines for actual examination. It can also be updated in real time or on a regular basis.

Also, each detailed registration possibility determination engine 400-1a to 400-1n may correspond to each clause of the law, and may store the result of the determination in relation to each clause of the law.

Trademark DB similar to registered trademark

It is possible to build a DB for trademarks similar to the registered trademarks, trademarks similar to the ones that have been extinguished for less than one year, and registered trademarks (collectively referred to as "priority trademarks"). The property of this DB is that data addition and deletion occur frequently.

The preceding trademark DB is made by processing a trademark-related publication issued by the JPO as a source. The JPO has various notices regarding trademarks, which are generated online or offline (CD-ROM, etc.) by electronic (digital) or printing methods, and can update the preceding trademark DB by collecting, extracting, or inputting them. have. The updated data can be transmitted to the person or system related to the implementation of the present invention, and the computer of the person who receives the data can update his DB.

Foreign registration possibility judgment system (400 / 1-1)

It is a system to find out whether a Korean trademark can be registered in a specific foreign country such as Japan. The essential structure of this system is very similar except for the system for checking the possibility of Korean trademark registration in Korea and the difference in the detailed composition due to the difference in registration requirements for each country. This goes through the following processors:

First, select a country to look for trademarks, products and registration possibilities. The system recognizes a country, and if there is a registration possibility determination system 400 / 1-1 for that country, it can query the system therein. Otherwise, information about trademarks and designated products is transmitted to a system that can determine the likelihood of registration in that country at a foreign or other location compatible with the system. In this case, if there is a uniform code for the designated product (classification code according to NICE classification), only the code can be provided. If the identity of the designated product is not maintained, the nearest designated product is designated as the designated product. The possibility of registration can be determined for the product. At this time, the registration possibility determination system (400 / 1-1) of the country may go through the process of processing the trademark (2-3).

Second, the registration possibility determination system 400 / 1-1 based on the country is inquired of the registration possibility. In some cases, the country-based registration likelihood determination system (400 / 1-1) reviews only some limited registration requirements, not all registration requirements. For example, a site in Japan (such as the Japan Patent Office) may only perform a similarity search for a trademark, and in this case, a similarity judgment may be made rather than a judgment on the whole possibility of registration. For this purpose, the process of processing a trademark (translating into Japanese, transliteration, etc.) may be presumed so that the system can be inputted into the system. At this time, if the country and Korea have the same requirements in the detailed registration requirements, some of the common registration requirements may be determined under Korean standards. This may be particularly useful if the country does not provide a registration requirement determination provided by the registration possibility determination system 400 / 1-1 in Korea.

Third, the result of the registration possibility is output, and optionally, the source data is translated as it is or translated into Korean. When the determination of the possibility of registration (including the likeness or not) is performed outside of the internal system, the information may be transmitted to the system and output to the user.

This can be particularly useful in the designation process of designated countries in international registration of trademarks under the Madrid Protocol. In other words, if there is minimal information about the possibility of registration in a particular country, the processing of part (especially the part of the application) or the entirety of the international application of the trademark (the application of the trademark under the Madrid Protocol to a foreign country) is carried out online. May be applied in the selection process of the designated country.

The similarity determination / registration determination process will be described below with reference to FIGS. 5 and 6.

First, there is a process of inputting a trademark / good (2-2) and processing of a trademark / good (2-3). In this regard, a trademark similarity judgment (5-1), a product similarity judgment (5-2), and a result of the judgment are prepared in a report (5-3).

Second, independent of the first process, it is possible to carry out the registration possibility determination process (5-4) and to prepare a registration possibility determination report (5-5).

Third, when the report is output (5-6), the report is transmitted and the application intention from the recipient (5-7) is received, the online electronic filing process (5-8) / (530) can proceed. have.

6 is a conceptual flowchart for a registerability determination processor.

First, there is input of a trademark / designated product (2-2), and processing for the trademark / product (2-3).

Second, a trademark / product is entered into the detailed registration possibility determination engine 11-1 for the first registration requirement, and the determination engine determines the detailed registration possibility determination rule DB (11-3) and the first registration for the first registration requirement. Using the detailed registration possibility judgment data DB (11-2) on the requirements, it is determined whether registration is possible by the first registration requirement (11-4), and a judgment basis report is prepared (11-5).

Second, perform the first decision process (11-6) above for final registration requirements from the second.

Third, the results of the detailed registration possibility determination engines for each registration requirement are synthesized (11-7), the comprehensive judgment on whether registration is possible (11-8), the judgment basis report preparation (11-9), and the results are output ( 11-10).

Reporting system

Concept and Necessity

The reporting system is the applicant information, the trademark information, the designated product information, and the registration possibility determination system (400 / 1-1) found by the applicant discovery system (510) to inform the applicant of the registration possibility of the trademark, and responds It refers to any system for the purpose of receiving a message.

Configuration of the reporting system

The reporting system consists of 1) a report preparation system 300-5, 2) a sending and receiving system, and 3) a trademark synthesis system.

Report Writing System (300-5)

The report generation system automatically synthesizes the report. The content of the report includes the following: 1) Applicant's label, Label, Designated Goods, and other bibliographic information; 2) Display of any analytical information regarding the possibility of registration; 3) Same or similar designated product for the same trademark. Information about two or more applicants, 4) information about competitors or competitors, 5) guidance, usage, or contact point for the online filing system, for example, the address of a website with an online filing system. Etc.) or 6) other relevant information.

In order to automatically synthesize a report, a typical DB report writing technique or an electronic catalog production technique can be used. 3) The method of presentation of the information part for two or more applicants for the same or similar designated goods for the same or similar trademarks of (1) may be: May be added (especially when sending out on paper, etc.), or 2) hyperlinking to a place where all the information is connected. 3) If the system accepts the intention that the applicant wants information about another applicant, the system may accept the intention and extract information about the other applicant and transmit it to the applicant.

Sending and receiving system

The sending and receiving system refers to a system that functions to send a completed report to the applicant and receive the applicant's response.

Offline shipping

Offline shipment refers to a form output on paper or other offline media to be sent to the user without the wired / wireless network 1000. Mailing is one of the simplest examples. To this end, the process includes obtaining, confirming and determining a contact point for the applicant, outputting a report, and sending a report.

Online shipping

The online delivery is the transmission through the wired / wireless network 1000 and is delivered to the applicant through electronic mail or other message delivery means.

Online filing system

Brand synthesis system

The trademark synthesis system refers to a system for producing a trademark using the network 1000 technology. In particular, it refers to a system that serves to enhance the possibility of registration by creating a new trademark by adding a discriminating means to a trademark which is judged to be low in registration, particularly a trademark which is judged to be inferior in discrimination.

In general, this system has a low possibility of registering a trademark that has only a technical mark or only a prominent geographical name, and thus, pays attention to the “only” and combines a new symbol, character, figure, three-dimensional shape, or color to acquire the overall identification power of the trademark. It helps to do it.

The trademark synthesis system goes through the following process.

First, the applicant's registration possibility of the trademark is judged, and if the possibility is below a certain standard, the applicant is sent any judgment information on the possibility of registration and information on various means for improving the identification power.

Secondly, the applicant agrees to use the means to enhance his or her discriminating ability online or offline.

In the case of on-line, it can go through the proposal of the system's identification factor and the applicant's selection and consent in real time or at a certain time. This proposal and selection process can be repeated several times. In this case, a graphic program (including a plug-in) may be used to show a combination of identification enhancing means for a trademark to be applied online.

Offer additional components of discernible trademarks, such as by mail, even when offline, and send responses to documents or other means expressing your choice and consent to those in charge of or related to the system. can do.

Online payment system

The Trademark Registration Application Process and Use of the Trademark Synthesis System Others In order to utilize the various systems of the present invention, charges can be combined for each process or for each system unit. Payment may be made by credit card, mobile phone payment, online remittance, or other common payment method commonly used on the Internet. In addition, an offline method (transfer and deposit confirmation) can also be used as a payment method.

Online filing system

The online application system refers to a system for applying for trademark registration online. This includes, in addition to the determination of trademarks and designated products, which are various elements for the application, information or information on the applicant, forms of power of attorney or other documents that correspond to statutory requirements for the application or any other submitted documents. This is necessary.

The online filing system receives the information sent or input from the user and processes, modifies, or re-enters it to fit the legal filing form.

Since the applicant's trademark is basically processed as an image file, when the applicant applies for a trademark consisting of letters or symbols only, the applicant may enter an input of the characters or symbols to automatically generate an image.

Through this process, bibliographic details, trademarks, designated products, and other documents to be submitted are determined, which can be printed out and filed with the Patent Office in an offline or online manner.

In case of filing with the Korean Intellectual Property Office, this information can be used for electronic filing.

The effects of the present invention are as follows.

First, trademark similarity judgment and product similarity judgment can be made to search whether there is a trademark similar to the prior application or registered trademark before application. By combining linguistic engineering judgments about similarity, the data on the basis of similarity judgment can be automatically identified, and the report can be automatically generated and presented to users. This makes it easy to write opinions and increase objectivity in similar judgment. In addition, time cost can be reduced by automating manual similarity judgments. In addition, the morphological analysis engine for similar judgment and other various analysis engines presented in this specification can be applied to trademarks of other countries, thereby enabling analysis and similar judgment of foreign trademarks. This will be particularly effective in designating designated countries under the Madrid Protocol system.

Second, it is possible to grasp the possibility of registering a trademark in advance so that not only the trademark industry but also the general public can find out the possibility of registering the trademark in advance. In addition, it is possible to eliminate the waste caused by the trademark registration application that does not meet the registration requirements, thereby increasing the efficiency of the application.

The effects of the invention of each detailed system are often mentioned in the concept and necessity sections of this specification.

1 is a conceptual diagram of a system and network configuration of the present invention.

2 is a flowchart illustrating a configuration and trademark similarity and product similarity determination procedure for a trademark similarity determination system and a product similarity determination system.

3 is an explanatory diagram of various DBs related to the present invention including a trademark DB.

4 is a flowchart illustrating a process of forming a trademark and product similarity judgment rule and updating the same.

5 is a flowchart of a similarity determination and a registration possibility determination process.

6 is a flowchart illustrating a registration possibility determination process.

7: is a block diagram of a registration possibility determination system.

Claims (1)

An online step of inputting an applicant's trademark (2-2) and an applicant's designated product input (2-2), automatically determining the possibility of registration of the inputted trademarks and products, and outputting a result of the registration possibility online; Method for determining the trademark registration possibility, comprising the step of automatically synthesizing the evidence report of the registration possibility