JP2009176117A - Trademark search system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a trademark search system that allows a user to obtain a sophisticated search result close to a result of determination executed in more actual decision on appeal or the like. <P>SOLUTION: The trademark search system includes: an appellation-information processing means 101 for processing appellations of search targets; an inspection-target set preparation means 102 for extracting trademark information having appellation information matched with the processed appellation information from an appellation database so as to make the trademark information as an inspection-target set; a sequence-definition imparting means 106 for imparting a sequence definition regarding the appellation information included in the inspection-target set and appellation information of the search target for each sound constituting the appellation; an inspection-range extraction means 103 for setting mismatched parts of sounds of the appellation information to be compared with each other in the inspection target to be an inspection range; and a difference calculation means 104 for calculating a difference on the basis of the sound sequence definition within the inspection range. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a trademark search system for determining the similarity of a trademark or a naming, in particular, the similarity determination of the title.

  To determine the possibility of registration of a trademark to be applied for trademark registration, use various databases and use the trademark designation (reading) as a search key. Searching and extracting items that match the above. This makes it possible to check whether the same or similar designation has already been filed or registered by another person with reference to the list of search results.

  However, it is difficult for a person to determine whether or not the name used as a search key is similar to other names in the search list. Therefore, it is preferable to have a system for determining similarity with names in the list after obtaining the search list.

  By the way, since the Patent Office has previously published trademark examination standards and disclosed certain similar standards therein, it is preferable that these standards can be used for the search and subsequent similarity determination. In addition, if it is possible to obtain and display the similar or dissimilar name for the reason why the two names are compared, it is possible to make a comparatively objective and persuasive similarity determination.

  In view of the above problem, a system is known that can determine the similarity between two names to be compared based on the JPO Trademark Examination Standard (see Patent Document 1).

JP 2001-222557 A

  The criteria set forth in the examination criteria for determining whether a trademark is similar to that of a trademark, such as whether the prefix is similar to a character string, especially consonants, are similar to the trademark structure. It is a general-purpose criterion that takes into account. However, in reality, there are quite a few cases where two trademarks that are judged to be similar according to the examination standards are judged to be dissimilar in judgment and judgment.

  The difference between this examination standard and actual judgment is based on, for example, phonological differences based on phonology, and similar judgments are made for such individual specific contrast trademarks in consideration of such information. Therefore, in the system described in Patent Document 1, sufficient similarity determination is impossible.

  Therefore, the technical problem to be solved by the present invention is to provide a trademark search system system capable of obtaining a high-level search result closer to a determination result made in an actual decision in addition to similarity determination based on examination criteria. Is to provide.

  In order to solve the above technical problem, the present invention provides a trademark search system having the following configuration.

According to the first aspect of the present invention, it includes a process of determining whether or not the names of two trademarks to be compared between the names included in the trademark information stored in the name database and the names to be searched are similar to each other. A trademark search system,
The name information of the trademark stored in the name database includes original name information derived from the structure of the trademark and processed name information processed from the original name information,
Name information processing means for processing the name of the search target and adding it as name information;
Inspection object set creation means for extracting, as the inspection object set, trademark information in which the original name information and processing name information of the trademark stored in the name database matches the name information of the search object processed by the name information processing means When,
For the original name information of the trademark information included in the set to be examined and the name information of the search object, sequence definition giving means for giving a phonological arrangement definition for each sound constituting the name,
A test range extraction means for comparing the sound of the name information of the test object and the original name information of the trademark information included in the test object set, and specifying a mismatched portion as the test range;
A difference calculating means for calculating a difference indicating a similarity between two names to be compared based on an array definition given to the sound within the examination range specified by the examination range extracting means; A trademark search system is provided.

  According to a second aspect of the present invention, there is provided the trademark search system according to the first aspect, wherein the array definition is a numerical value defined for the articulation method and the articulation location for each sound constituting the designation. To do.

According to a third aspect of the present invention, the name information processing means includes the original name call information included in the inspection target set and the name information related to the search target.
When long sounds, prompt sounds, subordinate sounds, and double vowels are included, naming information excluding these,
Naming information excluding these when long sounds, prompt sounds, long-sounding sounds and double vowels are included,
If the sounding conditions are met, the name information for the sounding,
If prompting sounds and special roaring sounds are included, naming information with capitalization of them,
Name information that unifies specific sounds, straightened stuttering, and unified muddy / semi-voiced sound,
Name information for differences and omissions depending on the number of sounds that make up the evidence,
The trademark search system according to the first or second aspect is provided, characterized in that at least one process name information is added.

  According to the fourth aspect of the present invention, the inspection range extraction means compares the original name call information included in the inspection object set and the sound position of the search target name information in common, Alternatively, the trademark search system according to any one of the first to third aspects is characterized in that the sound position of the name information to be searched is shifted by one character before and after the comparison, and the inconsistent portion is specified as the inspection range. I will provide a.

  According to the fifth aspect of the present invention, the difference calculation means calculates the number of differences based on the absolute value of the difference between the numerical values of the array definitions given for the sounds included in the examination range of the title to be compared. A trademark search system according to any one of the first to fourth aspects is provided.

  According to the present invention, it is possible to narrow down the target for calculating the number of differences indicating similarity by processing the name of the trademark information stored in the name database and the name of the search target to create a set of inspection targets. It is possible to speed up the processing. Further, the number of different points is configured by phonological arrangement definition, and is calculated based on the configuration of the sound of the mismatched portions of the two names to be compared, so that the difference between the names can be calculated more accurately. Therefore, it is possible to obtain an advanced search result that is close to the judgment result made in an actual trial decision or the like.

  Hereinafter, a trademark search apparatus according to an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a system configuration diagram of a trademark search system configured by a trademark search apparatus according to an embodiment of the present invention. The trademark search system 1 includes a network in which a trademark search server 2 and a plurality of operation terminals 3 are connected via the Internet IN.

  The trademark search server 2 is a server that performs a search process transmitted from the operation terminal 3 and transmits a search result to the operation terminal 3. In the present embodiment, a similar search call database that is a database used in the search process performed by the trademark search server 1 is configured to be provided in the server (see FIG. 7), but the trademark search server 2 It may be configured separately.

  The operation terminal 3 is a terminal used by a person who searches for a trademark using the trademark search system 100 according to the present embodiment, and is specifically configured by a general-purpose computer. A trademark searcher who is a user of the operation terminal 3 registered in advance in the trademark search server 1 accesses a specific URL using the operation terminal 3 and operates using the GUI to operate the trademark search server 1 on the web. The search result can be obtained by performing a search by operating and displaying on the own operation terminal 3 by means such as display or printer output.

  The trademark search system according to the present embodiment searches whether or not a trademark related to an inputted name of a search target is filed or registered based on information related to a name stored in a similar search name database. Alternatively, the degree of difference between the registered trademark and the trademark to be searched can be calculated as a numerical value. Before describing in detail the system configuration of the trademark search system according to the present embodiment, the trademark search procedure and information displayed by the trademark search system according to the present embodiment will be described.

  FIG. 2 is a flowchart of search processing as seen from the movement of the screen displayed on the operation terminal of the trademark search system according to the present embodiment.

  First, the searcher operates the operation terminal 3 to access a predetermined URL. Receiving access from the operation terminal 3, the search server 1 activates the web program and performs processing to display a login page on the operation terminal 3. On the login page (not shown), an input field for inputting a user ID and password is displayed. When the searcher inputs and transmits necessary information, the search server 1 receives the information and stores it in the customer database. Compare with the information you have. If the user ID and password are correct, the main menu is displayed. If the user ID and password are illegal, an error message is displayed and the user ID and password are requested to be input again.

  FIG. 3 is a diagram illustrating a screen display example of the main menu. The main menu is provided with six submenus. By selecting the name similar button 11, the full text search button 12, and the bibliographic search button 13, the trademark search program stored in the trademark search server 2 is started. Then, a trademark search menu for searching for a trademark is activated. The trademark search menu is configured to be able to search for trademarks using various search keys. For example, bibliographic information such as applicant and application date may be used as a search key, or a designated product / A service, a similar group code, or the like may be used as a search key. Also, a trademark as a mark may be used as a search key, or a trademark name may be searched as a character string.

  The trademark dictionary button 14 is a button for executing a trademark dictionary function disclosed in Japanese Patent Application No. 2007-277573 filed by the applicant. When a trademark is input, a name assigned to the trademark is searched. can do. The recording range button 15 is a button for displaying the recording range of data stored in the similar search name database of the trademark search server, and the usage details button 16 is used by the user specified by the user ID in the system. It is a button for displaying the usage details using.

  The book review search system according to the present embodiment can search for trademarks or registered trademarks that have been applied using various search keys as described above. The following is a feature of the trademark search system according to the present invention. A case where a name of a trademark is selected as a search key, that is, a case where the name similar button 11 is operated in the main menu shown in FIG. 3 will be described.

  FIG. 4A is a diagram showing a display example of a condition input screen of the trademark search system according to the present embodiment. In the condition input screen, a name input field 21, a similar group code input field 22, and an international classification input field 23 are provided as input fields for inputting search conditions. After entering the search conditions (# 1), search is performed. By operating the execution button 26, it is possible to search for a trademark that matches the input search condition.

  In the name input field 21, a trademark dictionary window 24 for searching for a name candidate using the above-described trademark dictionary function, and a name input window 25 for determining the name candidate displayed in the trademark dictionary window 24 as a search condition. Is provided.

  4B is an enlarged view of a range B in FIG. 4A. The trademark dictionary window 24 shown in FIG. 4B is a column in which candidate names assumed from the configuration of the trademark input in the trademark input column 24a are displayed. The trademark dictionary is used to extract a candidate for a name assumed from the configuration of the trademark when the name is not uniquely specified from the configuration of the trademark, for example, when the trademark is written in alphabet or Kanji. This is an input assist function. When the name extraction button 24b is operated in a state where the trademark configuration is input in the trademark input field 24a, the similar search name database in the search server 2 is searched and the candidate names assumed from the configuration of the trademark. Are displayed as a list (# 3). The specific configuration of the trademark dictionary processing system is disclosed in Japanese Patent Application No. 2007-277573 filed by the applicant.

  FIG. 4C is an enlarged view of a range C in FIG. 4A. A name input window 25 shown in FIG. 4C is a column for selecting and inputting a name as a search condition by appropriately operating the operation button 25a among the name candidates listed using the trademark dictionary (# 2). It is. In the name input window 25, not only the name candidates listed in the trademark dictionary window 24 may be selectively input, but also names as search conditions may be directly input.

  When the search execution button 26 is operated, if there is no search condition error (# 4), a trademark that satisfies the search condition is searched. Examples of the search condition error include a case where a title as a search condition is not input in the title input window 25. Note that the search conditions in the similar group code input field 22 and the international classification input field 23 can be arbitrarily entered. If no search condition is input in the respective fields, the search is executed using all classifications as criteria.

  FIG. 5 is a screen display example showing a part of the search result display screen in a state where search execution has been performed and search at the same level has been completed. The display of the number of results of the same level search, which is the result of the search at the same level (# 5), is displayed in the same search result column 28 for each name as the search condition 27, exact match, forward match, intermediate match , The backward match result is displayed. The search result at the same level is temporarily stored in the trademark search server, and is used to create a list display that is displayed when the list display button 32 is operated.

  In the screen display shown in FIG. 5, the similarity search can be further performed (# 7). The similarity search is executed by operating the similarity search field 29 or the similarity search button 31. The trademark search system according to the present embodiment can evaluate the similarity between the name of the search target and the search result trademark as a degree of difference. As will be described later in the processing, the degree of difference is obtained by calculating the degree of difference between the two trademarks as a numerical value in consideration of characteristics based on phonology for the two trademarks to be compared. In this embodiment, the degree of difference is set to be 0 (indicating that it is a trademark having the same designation), and the degree of difference increases as the numerical value increases.

  6A and 6B are screen display examples showing a part of the search result display screen in a state where the similar search is finished. The similarity search can be performed for each arbitrary designation as a search condition as shown in FIG. 6A, or the designation as a plurality of search conditions specified as shown in FIG. 6B can be performed at a time. In the example shown in FIG. 6A, the dissimilarity shown in the dissimilarity column 29b is appropriately set, and by operating the search button 29b, the similar trademark satisfying the dissimilarity set in the nominal difference is satisfied. The number of hits is simply displayed (# 8). Further, in the example shown in FIG. 6B, the similarity satisfying the condition of the dissimilarity set as the dissimilarity by operating the similar search button 31 for the name of the search target selected by checking the selection column 29a. The number of trademark hits is simply displayed (# 8). Note that the search result regarding the similar trademark used for the simple display is temporarily stored in the server for use in the merge process.

  The two merge buttons 30 are operated when displaying similar search results for the names that are a plurality of search targets by AND merging or OR merging.

  FIG. 7 is a screen display example showing a part of the list display screen displayed after the similarity search is completed. As shown in FIG. 7, the list display is displayed by operating the similar search result number display portion. It is possible to check and narrow down the search results not in the trademark unit but in the name unit excluding duplication of the same name (leaving only the necessary name). On the list display screen, the name 32 as a search result and the numerical value 31 of the degree of difference when the name is compared with the name being searched are displayed. The bibliographic search (13) can be continued for the application or registered trademark whose check column 30 is checked.

  Next, a specific system configuration of the search server for realizing the search operation will be described. FIG. 8 is a block diagram showing a functional configuration of the trademark search server. The trademark search server (hereinafter sometimes abbreviated as “search server”) includes the functional blocks shown in FIG. 8, and these functional blocks respectively execute the processes described below.

  As shown in FIG. 8, the trademark search server stores a similar search title database 100 (hereinafter, sometimes simply referred to as a title DB). The similar search title database 100 stores information on trademarks applied for or registered with the Patent Office. As information registered in the similar search name database 100, for example, data obtained by performing predetermined processing on the trademark arrangement / standardized SGML data issued by the Patent Office and adding search name information can be used. The name information for search stored in the similar search name database 100 is created by the name information processing means 101 in the following procedure based on the above-mentioned trademark arrangement / standardized SGML data, and stored in the similar search name database 100.

First, another name information is created according to the following rules based on the information on the name of the trademark arrangement / standardized SGML data issued by the JPO, and added to the search name information in the similar search name database 100. To do. Specifically, name information is added according to the following rules.
(1) If the pre-processing name information includes a long sound, a prompt sound, a sound following a long sound, and a double vowel, a name excluding these is additionally created. For example, when the pre-processing name is “Sister”, “Sister” is added as name information.
(2) If the pre-processing name information matches the sounding condition, a name that has been sounded is additionally created. For example, when the pre-processing name is “swing”, “singing” is added as name information.
(3) If the pre-processing name information includes prompting sounds and special stuttering sounds, additional names are created by capitalizing them. For example, “toe” is added as the name information when the pre-processing name is “tour”, and “net” is added when the name before processing is “net”.
(4) Unify the sound of the specific sound included in the pre-processing name information, make the stuttering sound straight, and unify the muddy / semi-voiced sound and add it as name information. For example, “wo” is replaced with “o”, “di” is replaced with “ji”, “va” is replaced with “ba”, and “fa” is replaced with “ha”.
(5) For the name information created in the above (1) to (4), name information for difference and omission is created according to the number of sounds constituting the name call. Specifically, in the case of the same search, the above-mentioned designation information is used as it is, but in the case of a similar search, a character string at an arbitrary position is blank according to the number of sounds (determined based on whether or not there are 7 or more sounds). Create missing name information with differences replaced with characters and any character strings deleted.

  An example in which the nominal information processing unit 101 processes nominal information according to the above rule will be described using a specific example. For example, when the pre-processing name information that is the trademark arrangement / standardized SGML data issued by the JPO is “Dussia Pitch”, “Dusia Pichi” is added according to the law (1) for deleting long sounds and prompt sounds. The "Dussia pitch" is added according to the law (2) for creating a stuttering character. In accordance with the law (3) for capitalizing prompt sounds and special stuttering, “Dousia pitch” is added. In accordance with the unification of the special sounds of Law (4), “Zussia Pitch”, “Zussia Pissi”, “Zussia Pissi”, “Dousia Hissushi” are added, and “Sussia Hissi”, “Sushihashi”, "Susahishi", "Toussiahitsushi" (the above four are used as naming information for the same sound search only), "Zussiahishi", "Zusabishiki", "Zoosabishiki", "Doushiabithi" (as naming information other than the same sound) are added. In accordance with the law (5) for creating sound differences and omissions, “○ -Shiabishi”, “Zu * Sabibishi” (one-tone difference), etc. (○ indicates any one character string), “XX Sheabishi”, “ “○○○ ABISI”, “○ OOSI ○ BISSI” (two-tone difference), “−siabiashi”, “Zushibiashi”, “Zooshibisshi” (one sound missing), and the like are created. Note that the processing name designation information for two-tone differences is created when the number of sounds created for the designation is 7 or more. For example, “Dussia Pitch” has 7 tones, so that designation information for two-tone differences is created. However, since “Zushiapishi” has 5 sounds, the name information of the difference between 2 sounds is not created.

  For each name information processed in accordance with the above rules, a name sound information storage array definition (hereinafter abbreviated as array definition) shown in FIG. 9 is given. The array definition is given by the array regular assignment means 106. The array definition is information obtained by digitizing the configuration of the name information based on the phonological characteristics, and information from 1 to 21 in FIG. 9 is initially added to the name DB. Specifically, 1 (oto) is information in which name information is represented in kana, and 2 (code) is information in which kana constituting the name information is converted into a 6-byte sound code. 4 (bs) of 3 (SS) is numerical information used for calculating the number of consonant and vowel missing insertion points in the comparison process described later. 5 (hs) to 8 (um) are information quantified with respect to the articulation method attached to each sound constituting the name information. 9 (kb) to 16 (kk) are information quantified with respect to the articulation location given to each sound constituting the name information.

  Each attribute value from 5 (hs) to 16 (kk) is determined based on the element classification table determined for each sound according to phonology. For example, the consonants of “B, B, B, B” are classified as voiced plosives, 0 for the articulation location “16 (kk) palate”, 4 for “6 (hs) closed”, and “6” An attribute value of 0 is assigned to (ms) friction, 0 to 7 (bn) nasal sound, 0 to 8 (um) utterance, and 4 to 3 (ss) consonant intensity. Also, for example, each vowel of “U, E, O” is classified as a narrow vowel, 0 in the articulation location “16 (kk) palate”, 0 in the articulation method “5 (hs) closed”, “6 ( ms) Friction ”is assigned 0,“ 7 (bn) nasal sound ”is 0,“ 8 (um) utterance ”is 1, and“ 4 (bs) vowel intensity ”is 2.

  In the array definition shown in FIG. 9, the attribute values from 22 (fmp) to 27 (rsl) are not stored in the nominal DB, and are subject to inspection created from the nominal DB in the search processing described later. After the set is created, it is given for each search process in the process of comparing with the name of the search condition.

  A search process for the trademark search system including the name DB 100 having the above configuration will be described. FIG. 10 shows a flowchart of search processing performed by the trademark search server of the trademark search system according to the present embodiment. The flowchart shown in FIG. 10 shows a process in the case where a similar search for trademarks as shown in FIG. 6A is performed, and when name information to be searched and information on the degree of difference are transmitted from the terminal device to the search server. It is a process performed. The search for the same trademark shown in FIG. 5 and the like may be performed by confirming the match between the name information stored in the name DB and the character string of the input information.

  In the condition input screen (see FIG. 4A), when a search condition (name and dissimilarity) is input and a similar search is executed, the name information is processed for the trademark that is the input search condition (# 21). .

  In the process of creating a set to be inspected, for the nominal information that is the search condition, the nominal information processing unit 101 adds the nominal information for search by the same processing as the processing of the nominal information performed in the nominal DB 100, and the array definition Is added.

(Inspection object set creation process)
Next, the inspection target set creation processing is performed by the inspection target set creation means 102 (# 22). In the process of creating the inspection target set, the name information to be searched is compared with the name information stored in the name DB 100, and the name information to be searched and the name information stored in the name DB 100 are found. Extract matching trademark information. The process of creating the inspection target set will be specifically described with a specific example. FIG. 11 is an explanatory diagram of the process of creating the inspection target set. As shown in FIG. 11, an example is given in which a trademark name “Dusia pitch” as a search target is input. Based on the inputted “Dusia pitch” designation, a processing designation 50 for search is created. Also, in the designation DB 100, a trademark having the designation “Zuushavic” is registered as application or registration information, and the machining designation information 51 processed based on the designation information “Zuushavic” is stored in the designation DB 100. Yes.

  The inspection target set creation unit 102 compares the processing name / call information 50 to be searched with the processing name / call information 51 stored in the name DB 100, and searches whether any one matches. As a result, the original name information of the trademark information having the matching name information (that is, “Zuschavic before processing”) is extracted from the name DB 100 and listed in the inspection object set.

  Hereinafter, the processed name information comparison process is performed for all the trademark information stored in the name DB 100, and when the comparison process ends, the inspection target set creation process ends.

(Inspection range confirmation process)
The trademark search server 2 performs an inspection range determination process following the inspection object set creation process (# 23). The inspection range determination process is executed by the inspection range determination means 103, and the inputted name information of the search object is compared with the original name information included in the inspection object set creation process, and each sound is compared with the other party. This is a process for determining the inspection range to be compared with which sound of the sound. The two names are compared for independent sounds other than long sounds, prompt sounds, and sounds similar to long sounds.

  FIG. 12 shows a flowchart of the examination range determination process. In the inspection range determination processing, the same position confirmation (# 231), shift position confirmation (# 232), and inspection range extraction (# 233) are performed for each piece of trademark information listed in the inspection object set.

  The same position confirmation (# 231) is a process for confirming whether or not the sounds of two pieces of name information to be compared are in the same position. FIG. 13 shows a schematic diagram of comparison of naming information in the same confirmation process. In FIG. 13, if the sound position of the name information of the name information of the inspection target set to be inspected and the name position of the name information of the search target to be inspected are at the same position, whether or not those sounds are common The state which compares is shown.

  The shift position confirmation (# 232) is a process for confirming whether or not the sounds are common when the sounds of the two pieces of name information to be compared are at positions shifted to ± 1 positions. 14A and 14B are schematic diagrams showing comparison of naming information in the shift position confirmation processing. In FIG. 14A, the sound when the first character's name position of the name information of the inspection target set to be inspected is shifted to the end of one sentence of the name information's name position which is the search condition to be inspected, That is, it shows a state in which it is compared whether or not it is common with the sound of the first character and the second character of the test subject calling sound. Further, FIG. 14B shows that the name position of the second character in the name information of the set of inspection targets to be inspected is shifted to the end of the sentence and the beginning of the sentence in the name information of the name information which is the search condition to be inspected. It shows a state in which it is compared whether the sound is the same as the sound of the first character and the third character of the test name calling sound.

  As a result of the same position confirmation (# 231) and shift position confirmation (# 232), in the case of the same sound, the comparison definition position (fmp) and comparison end position ( The sound position of the other party (name information to be searched) is input to (top), and 1 (determined) is set in the sound position determination state flag (ckp), and 1 (same) is set in the difference state flag (rsl).

  Next, the inspection range is extracted (# 233). In the extraction of the inspection range, adjacent sounds other than the combination in which the same sound is found are grouped as one group, and the inspection range is calculated. FIG. 15A shows a schematic diagram of the inspection range extraction process. FIG. 15A shows that the same symbols are the same call sounds. In FIG. 15A, the inspection side 1st to 3rd sound, the inspected side 1st and 2nd sound, and the inspection side 7th and 8th sound and the inspected side 6th to 8th groups are compared. That is, the fourth to sixth sound on the inspection side and the third to fifth sound on the inspection side are common and are not extracted as the inspection range.

  The inspection range extraction processing will be described with reference to FIGS. 15B and 15C with a specific example. As shown in FIG. 15B, in the name comparison between “Ciderville” and “Sideval”, the inspection range is different between “Dabi” and “Doba” in the third and fourth sounds, so the third and fourth sounds are the inspection range. Become. Further, in the example of FIG. 15C, comparison of designations of “Vitaflex” and “Biplex” is made, and the sound excluding the same sound position “Bi” and “Rex”, that is, “Tough” of the second and third sounds is the inspection range. It becomes.

(Difference calculation)
When the inspection range is extracted by the above process, a comparison sound is determined within the inspection range, and the number of differences (hereinafter abbreviated as differences) is calculated based on the sound within the inspection range (# 24- # 29). The difference calculation is executed by the difference calculation means 104. The difference is a quantification of the degree of phonological difference between the sounds specified by the two designations to be compared, and is calculated based on the sequence definition assigned to each designation. As shown in the flowchart of FIG. 10, the calculation of the difference is performed by a plurality of types of calculation methods, and the combination having the smallest difference (that is, the highest similarity) is determined (# 30). Hereinafter, a difference calculation method based on each calculation method will be described.

  As a method of calculating the difference, calculation is performed for all sounds (independent sound, distinction of sound that follows long sound) (# 24), calculation is based on the arrangement definition of only strong sound (independent sound) (# 26), When stuttering is performed, calculation is performed for all sounds (when stuttering is included in one of the two names to be compared) (# 28), and a total of 4 transposition calculations (# 29) Calculate with the street method. Hereinafter, the calculation method will be described.

  FIG. 16 is a diagram illustrating a specific example of the inspection range extracted by the inspection range extraction process. In the example of FIG. 16, the two names to be compared are “Fantasick” and “Fan Pacific”, and “Tashi” and “Pacific” enclosed within the frame in the figure are the inspection range. Differences are calculated based on the array definition values to which the sounds are assigned in the following 11 patterns, while shifting the positions based on the order of the sounds in the inspection range.

The calculation pattern of the arrangement order within the inspection range is as follows.
(1) Arrangement position of test title and test title is the same position (2) Arrangement position when test title is shifted backward by one character (3) Arrangement position when test title is shifted backward by one character (4 ) Arrangement position when the test title is shifted two characters later (5) Arrangement position when the test title is shifted two characters backward (6) The first note is the same position, the second and subsequent notes are the test title 1 Arrangement position when the character is shifted backward (7) The first note is the same position, and after the second tone, the arrangement position when the test title is shifted backward by one character (8) The first note is the test title with one character Arrangement position when the test name is shifted backward by one character after the second sound and after the second sound (9) The first sound shifts the test name one character behind, and after the second sound, the test name is one character after the second sound Arrangement position when shifted backward (10) The first note shifts the name of the subject to be inspected by one character Arrangement position when the test name is shifted backward by one character after the second sound (11) The first sound is shifted backward by one character for the first sound, and the test title is shifted further by one character after the second sound Array position

  The arrangement sequence pattern will be described with reference to the drawings. FIG. 17A is a diagram illustrating a case where a difference is calculated using the pattern (1). As shown in FIG. 17A, in the pattern (1), the sounds at the same position within the examination range are respectively compared, and the difference is calculated based on the array definition. In the example of FIG. 17A, the test name is 3 tones and the name to be inspected is 4 tones, so the fourth sound is inserted using the consonant intensity 3 (ss) or vowel intensity 4 (bs) in the array definition. Is calculated.

  FIG. 17B is a diagram illustrating a case where a difference is calculated using the pattern (2). As shown in FIG. 17B, in the pattern (2), the test title is shifted by one character, and the difference is calculated. Since there is no opponent sound for the first sound of the test name call, the missing point is calculated, and for the third and fourth sounds of the test name call, the insertion point is calculated because there is no other sound.

  FIG. 17C is a diagram illustrating a case where a difference is calculated using the pattern (3). As shown in FIG. 17C, in the pattern (3), the name of the test subject is shifted by one character, and the difference is calculated. Since there is no opponent sound for the first sound of the name to be inspected, the insertion point is calculated.

  FIG. 17D is a diagram illustrating a case where a difference is calculated using the pattern (4). As shown in FIG. 17D, in the pattern (4), the inspection title is shifted by two characters and the difference is calculated. For the first and second sounds of the test name call, the missing point is calculated because there is no other sound, and for the fourth and fifth characters of the test name, the insertion point is calculated because there is no other sound.

  FIG. 17E is a diagram illustrating a case where a difference is calculated using the pattern (5). As shown in FIG. 17E, in the pattern (5), the inspected title is shifted by two characters and the difference is calculated. The insertion point is calculated for the first and second sounds of the name to be inspected because there is no other party sound, and the missing point is calculated for the fourth and fifth characters of the name of the inspection name because there is no other party sound.

  FIG. 17F is a diagram illustrating a case where a difference is calculated using the pattern (6). As shown in FIG. 17F, in the pattern of (6), the first sound is the same position, and the inspection name after the second sound is shifted by one character, and the difference is calculated. That is, the third sound of the test title is compared with the second sound of the test title. Since there is no opponent sound for the second sound of the test name call, a missing point is calculated, and for the fifth character of the test name call, an insertion point is calculated because there is no other sound.

  FIG. 17G is a diagram illustrating a case where a difference is calculated using the pattern (7). As shown in FIG. 17G, in the pattern (7), the first sound is at the same position, the inspected names after the second sound are shifted by one character, and the difference is calculated. That is, the third sound of the test title is compared with the second sound of the test title. The insertion point is calculated for the second sound of the name to be inspected because there is no other party sound, and the missing point is calculated for the fifth character of the name to be inspected because there is no other party sound.

  FIG. 17H is a diagram illustrating a case where a difference is calculated using the pattern (8). As shown in FIG. 17H, in the pattern (8), the difference is calculated by shifting the test title call by one character for the first sound and by shifting the test title call for the second and subsequent sounds by one character. That is, the second sound of the test name is compared with the first sound of the test name, the fourth sound of the test name and the second sound of the test name. For the first and third sounds of the test name, the missing point is calculated because there is no other party sound, and for the fourth and fifth characters of the test name, the insertion point is calculated because there is no other party sound.

  FIG. 17I is a diagram illustrating a case where a difference is calculated using the pattern (9). As shown in FIG. 17I, in the pattern (9), the test title is shifted by one character for the first sound, and the test title is shifted by one character for the third and subsequent sounds. That is, the third and subsequent sounds are inspected at the same position. Since there is no opponent sound for the first sound of the test name call, the missing point is calculated, and for the second character of the test name call, the insertion point is calculated because there is no other sound.

  FIG. 17J is a diagram illustrating a case where a difference is calculated using the pattern (10). As shown in FIG. 17H, in the pattern (10), the test title is shifted by one character for the first sound, and the test title is shifted by one character for the third and subsequent sounds. That is, the third and subsequent sounds are inspected at the same position. Since there is no opponent sound for the second sound of the test title call, the missing point is calculated, and for the first character of the test title call, the insertion point is calculated because the other party sound does not exist.

  FIG. 17K is a diagram illustrating a case where a difference is calculated using the pattern (11). As shown in FIG. 17H, in the pattern (8), the difference is calculated by shifting the test title call by one character for the first sound and by shifting the test title call for the second and subsequent sounds by one character. That is, the second sound of the test name is compared with the first sound of the test name, the fourth sound of the test name and the second sound of the test name. For the first and third sounds of the test name, the missing point is calculated because there is no other party sound, and for the fourth and fifth characters of the test name, the insertion point is calculated because there is no other party sound.

  Next, a specific calculation method for the difference will be described. As described above, the array definition attached to each designation defines the score obtained by quantifying the sound intensity for each element of intensity, consonant, and vowel shown in FIG. A description of the specific numerical values of each element is omitted. Then, the numerical values attached to these array definitions are multiplied by the weights of the different elements to calculate the difference, the number of missing points, and the number of insertion points. A specific calculation method will be described with reference to FIGS. 19A and 19B.

  FIG. 19A is a diagram for describing a method for calculating a difference when the sounds to be compared are “kyo” and “na”. A specific numerical value of the array definition defined for Kyo and Na in FIG. 19A is taken for each item of the array definition, and the weighting parameter is multiplied. The difference between “Kyo” and “Na” is 5.8, which is the sum of the weighted multiplication values calculated for each item of the array definition.

  FIG. 19B is a diagram illustrating a method for calculating an insertion point / missing point. In the example of FIG. 19B, a case where the sound of “B” is inserted or missing will be described. Since the consonant intensity attached to the sound “B” is 4 and the vowel intensity is 3, 6.6, which is the sum of numerical values obtained by multiplying these by the weighting parameter, is calculated as the insertion point / missing point.

  As described above, the numerical value for which the difference is calculated for all sounds within the inspection range is the basic score of the difference.

In the present embodiment, the basic points of the differences calculated as described above are adjusted based on the following criteria. In addition, description is abbreviate | omitted about the specific numerical value about how much a difference point is reduced.
(1) Adjustment by position, that is, adjustment is performed depending on whether the inspection range is the beginning or the end.
(2) Adjustment by sound type Weak sound is a sound that follows long sound / promotion sound / long sound (same vowel, ray, ou), repellent sound, double vowel (eye, au, ae, ui, ou). Reduce differences. In addition, the number of differences is reduced for a specific combination “Munun”.
(3) Adjustment for the presence or absence of weak sounds Reduce the differences in the missing and inserted words and endings of the sounds of “su, zu, le, te, de, nu, mu, n”.
(4) Adjustment by stuttering (5) Adjustment for difference in prompting sound (6) Adjustment for difference in sound repellent (7) Adjustment for difference in long sound (8) Adjustment by context (9) Adjustment for missing insertion

  The above differences are calculated for the above 11 patterns, that is, a total of 33 patterns in total for all sounds, strong sounds, and stuttering.

  Next, the difference by the transposition inspection is calculated (# 29). As shown in FIG. 20, the calculation of the difference due to transposition is an inspection that takes into account the change of sound.

  In the transposition inspection, as shown in FIG. 21A, an operation is performed to check whether or not the character strings in the inspection range of the inspection title are common, while shifting the position. Then, as shown in FIG. 21B, when the first sound of the test name call and the second sound of the test name call are combined, the second sound of the test name call is the leading sound for the unbound sound in the test name call. Check from the end to the end.

  In this way, the inspection is performed for the presence of connection with all sounds of the inspection title. FIG. 21C shows the combination of the first sound of the test name and the second sound of the test name, the second sound of the test name and the first sound of the name of the test name, the third sound of the test name and the fourth sound of the name of the test name In this state, the fourth sound of the test title and the third sound of the test title are confirmed.

  When the test for the presence of connection is completed for all the sounds of the test name as described above, as shown in FIG. 21D, the insertion point and the missing point are calculated for the sound for which no connection exists, and this score is calculated. It is a difference.

  After calculating the difference between the above four calculation methods, it is adopted as the difference between the two designations for comparing the minimum difference (# 30).

(Display on screen)
The processing up to the above difference calculation is performed for all the names included in the set to be examined, and when the calculation for all the names is completed, the screen display means 105 displays the similar search results for the name information related to the search object. Extraction is performed in the range of the set difference shown in FIG. 6A or 6B, and the number of extractions is displayed on the screen.

  As described above, according to the search system according to the present embodiment, based on the sequence definition defined by the phonological characteristics, the difference in sound within the inspection range of the two nominal information to be compared is regarded as a difference. It is possible to calculate and to determine the similarity of names based on phonology.

  In addition, this invention is not limited to the said embodiment, It can implement in another various aspect.

1 is a system configuration diagram of a trademark search system including a trademark search device according to an embodiment of the present invention. It is a flowchart of the search process seen from the motion of the screen displayed on the operation terminal of the trademark search system of FIG. It is a figure which shows the example of a screen display of a main menu. It is a figure which shows the example of a display of the condition input screen of the trademark search system concerning this embodiment. It is an enlarged view of the range of B of FIG. 4A. It is an enlarged view of the range of C of FIG. 4A. It is a screen display example which shows a part of search result display screen of the state where search execution was performed and the search of the same level was completed. It is a screen display example which shows a part of search result display screen of the state which the similar search was complete | finished. It is another screen display example which shows a part of search result display screen of the state which the similar search was complete | finished. It is a screen display example which shows a part of list display screen. The block diagram which shows the functional structure of a trademark search server is shown. It is a structural example of an array definition. It is a flowchart of the search process which the trademark search server of a trademark search system performs. It is explanatory drawing of the process of preparation of a test object set. It is a flowchart of a test | inspection range determination process. It is a schematic diagram of the comparison of naming information in the same confirmation process. It is a schematic diagram of the comparison of naming information in a shift position confirmation process. It is a schematic diagram of the comparison of naming information in a shift position confirmation process. It is a schematic diagram of the process of test | inspection range extraction. It is a figure which shows a specific example about the process of test | inspection range extraction. It is a figure which shows a specific example about the process of test | inspection range extraction. It is a figure which shows the specific example of the test | inspection range extracted by the process of extraction of a test | inspection range. It is a figure explaining the case where a difference is calculated with the pattern of (1) of difference calculation. It is a figure explaining the case where a difference is calculated with the pattern of (2) of difference calculation. It is a figure explaining the case where a difference is calculated with the pattern of (3) of difference calculation. It is a figure explaining the case where a difference is calculated with the pattern of (4) of difference calculation. It is a figure explaining the case where a difference is calculated with the pattern of (5) of difference calculation. It is a figure explaining the case where a difference is calculated with the pattern of (6) of difference calculation. It is a figure explaining the case where a difference is calculated with the pattern of (7) of difference calculation. It is a figure explaining the case where a difference is calculated with the pattern of (8) of difference calculation. It is a figure explaining the case where a difference is calculated with the pattern of (9) of difference calculation. It is a figure explaining the case where a difference is calculated with the pattern of (10) of difference calculation. It is a figure explaining the case where a difference is calculated with the pattern of (11) of difference calculation. It is a figure explaining the element of an array definition. It is a figure explaining the calculation method of the difference in case the sounds to be compared are "Kyo" and "Na". It is a figure explaining the calculation method of an insertion point and a missing point. It is explanatory drawing about the inspection method of transposition inspection. In a transposition inspection, it is a figure explaining the process of the step which confirms the connection of the 1st sound of a test title. In a transposition inspection, it is a figure explaining the process of the step which confirms the connection of the 2nd sound of an inspection title. It is a figure explaining the process of the step which confirms the connection of the 4th sound of a test name call in a transposition test | inspection. It is a figure explaining the process of the step which completes the confirmation of the connection of all the sounds of a test title in a transposition test, and calculates a difference.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Trademark search system 2 Search server 3 Operation terminal 100 Name database 101 Name information processing means 102 Inspection object set preparation means 103 Inspection range extraction means 104 Difference calculation means 105 Screen display means 110 Transmission / reception part

Claims (5)

A trademark search system including a process for determining whether or not a title of two trademarks to be compared between a title included in trademark information stored in a title database and a title to be searched is similar to each other,
The name information of the trademark stored in the name database includes original name information derived from the structure of the trademark and processed name information processed from the original name information,
Name information processing means for processing the name of the search target and adding it as name information;
Inspection object set creation means for extracting, as an inspection object set, trademark information in which the original name and processing name information of the trademark stored in the name database matches the name information of the search object processed by the name information processing means. When,
For the original name information of the trademark information included in the set to be examined and the name information of the search object, sequence definition giving means for giving a phonological arrangement definition for each sound constituting the name,
A test range extraction means for comparing the sound of the name information of the test object and the original name information of the trademark information included in the test object set, and specifying a mismatched portion as the test range;
A difference calculating means for calculating a difference indicating a similarity between two names to be compared based on an array definition given to the sound within the examination range specified by the examination range extracting means; A trademark search system.   The trademark search system according to claim 1, wherein the array definition is a numerical value defined with respect to an articulation method and an articulation location for each sound constituting the designation. The name information processing means, for the original name information included in the inspection object set and the name information related to the search object,
Naming information excluding these when long sounds, prompt sounds, long-sounding sounds and double vowels are included,
If the sounding conditions are met, the name information for the sounding,
If prompting sounds and special roaring sounds are included, naming information with capitalization of them,
Name information that unifies specific sounds, straightened stuttering, and unified muddy / semi-voiced sound,
Name information for differences and omissions depending on the number of sounds that make up the evidence,
The trademark retrieval system according to claim 1, wherein at least one process name information is added.   The inspection range extraction means compares the sound name of the original call information included in the inspection target set and the sound position of the search target name information, and then compares the original call information or the sound position of the search target name information. 4. The trademark search system according to claim 1, wherein a comparison is made by shifting one character at a time before and after, and an inconsistent portion is specified as an inspection range.   The difference calculation means calculates the number of differences based on the absolute value of the difference between the numerical values of the array definitions assigned to the sounds included in the examination range of the designation to be compared. 5. The trademark search system according to any one of 4.

Images