JP4614568B2 - Code search processing system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a code retrieval processing system that retrieves a code assigned from a name such as a financial institution name to which an individual code is assigned.
[0002]
[Prior art]
In recent years, for example, when applying for a credit card, an application form is often provided with an automatic transfer entry field such as a membership fee. In addition to the application form, it may be prepared as a dedicated deposit account transfer request form (automatic transfer use application form). The applicant fills in the name of the financial institution (including the store name of the branch, etc.) in this automatic transfer entry field, and the applicant does not enter in the entry field of the financial institution code (store code). It is common that the code of the financial institution is given later on the merchant side. For this reason, it is desired to reduce paperwork and improve efficiency when assigning this financial institution code.
[0003]
Conventionally, when a predetermined membership fee is automatically transferred from a financial institution's account, a deposit account transfer request form (automatic transfer use application form) is provided in an application form such as a credit card, or as a single form Prepared. The applicant, when specifying a financial institution in the deposit account transfer request form (automatic transfer use application form), fills in at least the name, branch name, deposit type, and account number, and passes them to the corresponding trader.
[0004]
On the merchant side, the operator refers to the financial institution code (store code) from the name of the financial institution (including store name) entered in the requested deposit account transfer request form (automatic transfer application form). It is common to do the work of granting.
[0005]
[Problems to be solved by the invention]
However, giving the financial institution code (store code) while looking at the deposit account transfer request form (automatic transfer use application form) requires time, especially when the processing is large, There is a problem that it is not efficient. These problems are not limited to the case where a financial institution code is assigned, but the same problem occurs when an individual code is assigned from a name entered in a form.
[0006]
Accordingly, the present invention has been made in view of the above problems, and an object of the present invention is to provide a code search processing system that improves the efficiency of office processing when an individual code is assigned from the name of an entry.
[0007]
[Means for Solving the Problems]
  In order to solve the above-mentioned problem, in the invention of claim 1, the organization name including at least a predetermined organization name entered in a predetermined form and including a predetermined number of lower names belonging to the organization is read as image data. Based on the recognized character code data recognized as a character code string, the corresponding predetermined number assigned to the organization name from the organization code table by recognizing the type of the organization name and the unique name of the organization name excluding the type Code search processing system that obtains the organization name code including the organization sub-code, for the recognition character code data indicating the organization name of the target,About the beginning or end of the recognized character code dataMatching is performed with a predetermined number of type-name character codes, a predetermined number of type names are specified, and the remaining part of the specified type name among the group names is referred to the group code table as a unique name. The unique name is obtained by sequentially matching with the character code of the unique name obtained with the specified category name, and the organization name corresponding to the degree of matching of the identified category name and the entire organization name of the unique name is predetermined. Matching processing means for performing a process for specifying the number, code extraction means for extracting a predetermined number of corresponding group codes from the group code table based on the specified number of group names specified, and changeable by external input, Display processing means for performing processing for displaying the specified number of organization names and the corresponding organization code.
[0009]
  Claim2~4In the present invention, “the group code table is created by associating a predetermined number of group names related to the predetermined group and a predetermined number of group codes corresponding to the predetermined group based on the given predetermined group information”. Configuration,
  "In response to an external confirmation signal for display in the display processing means, the data processing device includes an output file creation means for creating and outputting a data file of the confirmed organization name and organization code.
  “The display processing means includes a display button for displaying image data of the target organization name on a screen that displays the specified predetermined number of organization names and corresponding organization codes”.
[0010]
  In this way, for the recognition character code data of the organization name including the name belonging to the lower order entered in the predetermined form,Predetermined for either the beginning or end of the recognized character code dataMatching with the category name (character code) is performed to identify the category name, and the unique name corresponding to the identified category name is matched with the character code of the unique name obtained from the organization code table for the unique name. Identifying a predetermined number of group names according to the degree of matching of the specified type name and unique name, and extracting a group code including a predetermined number of group sub-codes of the specified number of group names from the table Process to be displayed. In other words, since the type name and unique name are specified from the organization name part of the entry, it is possible to extract the individual code of the organization given to the organization, and improve the efficiency of the paperwork when giving the individual code. It is possible to plan.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. In this embodiment, as an example, the case of searching for a code of a financial institution name that is an organization name is shown. The present invention can also be applied to a case where a code is searched for.
[0012]
FIG. 1 is a block diagram of a code search processing system according to an embodiment of the present invention. FIG. 1A is a block diagram of the overall construction of the code search processing system of the present invention, and FIG. 1B is a block diagram of the configuration of the code search processing system. In FIG. 1A, a code search processing system 11 is constructed by a code search processing program in a computer 12, and the computer 12 is connected to at least an input means 13 such as a keyboard and a mouse, and a CRT or the like. The display means 14 is connected.
[0013]
The computer 12 is configured to extract data from the financial institution information DB 15 (which may be an external storage medium) by the code search processing system 11, and outputs the financial institution name code created by the code search processing system 11 to the output file 16. Are output (output to an internal storage medium and an external storage medium). In addition, recognition data recognized by a character code in an external data recognition device 17 (which may be in the computer 12) to be described later is taken in by a predetermined data input means.
[0014]
On the other hand, the data recognizing device 17 creates recognition data 19 recognized by the character code by the data reading / recognizing means 18. For example, a deposit account transfer for automatically transferring from a bank account when applying for a credit card or the like. An image of the written form including the request form (automatic transfer application form) or the filled-in part of the application form 20 (here, all filled-in parts) for each defined area. The data is read and recognized as a character code string of each character code such as Shift JIS, and the recognition data 19 is created.
[0015]
In this case, image data obtained by cutting out the portion in which the name of the relevant financial institution is cut out is stored in the image data 19A as a file, for example, in association with an ID (identifier) number given in advance to the application form. In other words, in the present embodiment, all of the entry parts of the application form 20 are recognized character data, and the area of the part where the name of the financial institution is entered in the code search processing system 11 is specified by a set value as described later. A character code string of a financial institution name is extracted, and code search matching processing is performed. Note that the data recognition device 17 may read only the part in which the name of the financial institution is entered as image data and use it as recognized character code data.
[0016]
As shown in FIG. 1B, the code search processing system 11 includes a data input / output means 21, a process control means 22, a table creation means 23, a data extraction means 24, a matching processing means 25 (type name processing means 25A, unique Name processing means 25B), code extraction means 26, display processing means 27, and output file creation means 28 are provided as appropriate. The data input / output means 21 is a data input / output interface, and each of the set values for designating the area in which the name of the financial institution is entered in the recognized character code data from the input means 13 and the financial institution information DB 15. Data of a financial institution name (including a lower store name) and a financial institution code (including a lower store code) are input, and data is exchanged with processing means (not shown) on the computer 12 side.
[0017]
The processing control means 22 controls the entire processing for specifying a financial institution code (store code) by matching processing. The table creation processing unit 23 creates a financial institution unique name table in which a financial institution type name code, a financial institution name, a financial institution code, and remarks are associated with each other from the financial institution information DB 15 and stores a lower store type name. A store unique name table in which a code, a store name, a store code, and remarks are appropriately associated is created. A group code table is constituted by the financial institution unique name table and the store unique name table. Further, the data extracting means 24 extracts a recognized character code string of the portion of the target financial institution unique name (store unique name) from the recognition data 19 based on the set value and makes it a target of matching processing.
[0018]
Although the details of the matching processing unit 25 will be described later, the type name processing unit 25A does not accept the top of the recognized character code data indicating the name of the target financial institution for the application form 20A in FIG. A matching process is performed on the end with a predetermined number of character codes of a predetermined number of type names (described in FIG. 3), and a predetermined number of type names are specified. In addition, for the application form 20B in FIG. 2B described later, the type name is specified from the recognized character code data of the part indicating the type of the target financial institution name among the parts specifying the predetermined financial institution name. To do.
[0019]
Further, the unique name processing means 25B, for the application form 20A in FIG. 2A described later, uses the remaining part of the specified type name among the financial institution names as a proper name as a financial institution code table (financial institution). A unique name is identified by sequentially matching with the character code of the unique name obtained by the type name identified with reference to the unique name table, and the application form 20B in FIG. Based on the recognized character code data indicating the unique name part of the institution name, the unique name is obtained by sequentially matching the character code of the unique name obtained with the type name specified by referring to the financial institution code table. A process of specifying a predetermined number of financial institution names corresponding to the degree of matching of the entire financial institution name of the type name and the unique name is performed.
[0020]
Here, the matching process is a matching algorithm by a program. In specifying the store name, a predetermined number of similar store types and store unique names are specified by replacing the financial institution name with the store name.
[0021]
The code extracting means 26 extracts a predetermined number of corresponding financial institution codes (store codes) from the financial institution unique name table (store unique name table) based on the specified number of financial institution names (store names). The display processing unit 27 can be changed by an external input from the input unit 13 and displays the specified predetermined number of financial institution names (store names) and corresponding financial institution codes (store codes) on the display unit 14. Process. Then, the output file creation means 28 responds to the confirmation signal from the input means 13 on the display screen in the display processing means 27 in accordance with the confirmed financial institution name (store name) and financial institution code (store code). A data file (output file 16) is created and output.
[0022]
Here, FIG. 2 shows an explanatory diagram of a form book which is one object of the code search process. FIG. 2A shows a financial institution entry portion of a deposit account transfer request form (automatic transfer use application form) which is a form as an application form 20A. “Global Credit Union” is entered, and “Tokyo Chiyoda Branch”, for example, is entered in the box of the “Branch Name” entry field as the subordinate store name. In addition, the account type and account number are also entered in the box in the “deposit type” entry field. Then, leave the “Financial institution code” and “Branch code” fields blank, and accept the deposit account transfer request form (automatic transfer use application form). The code is automatically given by the code search processing system 11.
[0023]
Further, FIG. 2B is a lower order in which, for example, “Earth Bank” is entered in the box of the “Name” entry field in the application form 20B, and the number representing the type name bank is marked with “O”. This shows a case where, for example, “Tokyo” is entered in the box of the “branch name” entry column as the store name of the type name, and the number representing the branch is marked with “◯”. In this case, a correspondence table of type names corresponding to the type of code data to be marked with “◯” is prepared in advance (described later).
[0024]
FIG. 3 is an explanatory diagram showing an example of the types of recognition type names in the code search process. FIG. 3A is a diagram in which “financial institution type name code”, “abbreviation”, and “attribute” are associated with a “type name” of a predetermined type for specifying the type name of the financial institution name. These are included, for example, in a program that performs the matching algorithm (may be a separate table). Here, “Bank”, “Credit Bank”, “Credit Union”, “Labor Bank”, “Credit Agricultural Cooperative”, “Agricultural Cooperative”, “Dairy Agricultural Cooperative”, “Credit Fishery Cooperative”, It is indicated as an attribute whether these type names are added to the head or tail of the financial institution unique name. The financial institution type name code “00” indicates “no type name”.
[0025]
FIG. 3B shows an example in which a “store type name code” is associated with a “type name” of a predetermined type for specifying a type name of a store name, and is included in a program that similarly performs a matching algorithm. (May be another table). Here, “branch”, “branch”, “agency”, and “branch office” are generally attached to the end of the store unique name, and therefore all the attributes at the end are associated. The store type name code “00” indicates “no type name”.
[0026]
Next, FIG. 4 shows an explanatory diagram of the recognition pattern of the type name and the unique name in FIG. FIG. 4A shows a case where the type name or proper name of the recognized character code data is replaced with one of the regular type name or proper name by another character. On the other hand, it is a pattern in which one character is replaced with another character at each location. In addition, when a regular name is one character, it does not determine with replacement.
[0027]
FIG. 4B shows a pattern when the type name or proper name of the recognized character code data is missing (when not recognized) by one character from the normal type name or proper name. FIG. 4C shows that the type name or proper name of the recognized character code data is mixed by one character with respect to the normal type name or proper name (for example, when the kanji bias and structure are recognized as different characters). ).
[0028]
That is, the matching process of the type name and the unique name is sequentially performed in all patterns of FIGS. 4 (A) to 4 (C).
[0029]
Next, FIG. 5 shows an explanatory diagram of an example screen display of the code search result. FIG. 5A shows a code in the case where, for example, a regular (FIG. 2) financial institution name “Global Credit Union”, the character “Shin” in the type name of the recognized character code data is replaced with one of the characters. It is a screen of a search processing result. Here, in the “financial institution” area, the name “replacement for the earth” is displayed, and the operator can change the name of the financial institution and the financial institution code (the name of the financial institution with a high probability). And financial institution code are displayed). In addition, a “search” button is provided to display a window in which another financial institution name and financial institution code can be selected for the input changeable area.
[0030]
The search result screen displays an area of “conversion result”, and displays a list of “reference value” and “probability list”. Here, the “reference value” indicates the reference of the matching rate when selecting the character matching level (in this case, the matching rate) with respect to the recognized character code data, and can be changed freely by the “set” button. You can set the value. For example, the criterion value “0.9” is a strict judgment, the criterion value “0.7” is a normal judgment, and the criterion value “0.5” is a tolerance judgment to judge the validity of matching. Even when the matching rate of the first item is larger than the reference value, it is determined that there is a problem with the validity of matching when the difference from the matching rate of the second item is “0.1” or less.
[0031]
The item “probability list” displays the name of the financial institution in which the matching rate in the matching between the type name and the unique name of the financial institution is selected and determined according to the reference value. That is, since the result of the matching process is not necessarily one, the list is displayed in order from, for example, the one with the highest matching rate (probability). The matching rate determination process in the matching process will be described with reference to a flowchart described later.
[0032]
In the search result screen, an “OK” button and a “Cancel” button are provided, and the operator confirms that the financial institution name and the financial institution code displayed in the area where the input can be changed in the “financial institution” area are correct. When the determination is made and the “OK” button is clicked, it is input to the code search processing system 11 as a confirmation signal. When the “Cancel” button is clicked, the name of the financial institution and the financial code cannot be determined, and the search processing for the financial institution code of the recognized character code data is terminated.
[0033]
FIG. 5B is different from FIG. 5A in that, for example, for the regular (FIG. 2) financial institution name “Global Credit Union”, any of the characters “sphere” in the proper name of the recognized character code data is This is a screen of the code search processing result when the character is replaced with the character of the display screen, and the configuration content of the display screen is the same as in FIG. The screen of the code search process result in the store name matching process is the same.
[0034]
Next, FIGS. 6 and 7 show flowcharts of the entire code search processing according to the present invention. In FIG. 6, first, a code search processing program is started in the computer 12 to display a screen for processing selection (system construction, code conversion processing, end) (step (S) 1). When the input signal from the outside is an end signal (S2), the process is terminated. In the case of an input signal for system construction (S3), construction of the code retrieval processing system 11 is executed (S4). When this system is constructed, the table creating means 23 extracts the financial institution unique name (financial institution code) and the store unique name (store code) from the financial institution information DB 15, and the financial institution unique name table and the store specific, respectively. A name table is created and stored in a predetermined storage area (for example, an internal storage area such as a hard disk of the computer 12) (S5), and an input signal for code conversion processing is awaited (S3).
[0035]
Subsequently, when an input signal for code conversion processing is input (S3), settings for extracting the target character code data portion for the input recognized character data (setting values input from the input means 13) are made. (S6), the corresponding character code data is obtained from the recognition data 19 provided by the data extraction means 24 (S7), and based on the extraction setting, first of all, the corresponding character code data is converted into the name of the financial institution. The corresponding character code string is extracted (S8). Here, in the case of the application form 20A, the character code string of the financial institution name including the type name, and in the case of the application form 20B, the character code string of the name of the financial institution having only the unique name and the type are specified. This is a character code string of code data (portion marked with “◯”). In addition, the said setting considers the case where the forms of the said application forms 20A and 20B differ, respectively.
[0036]
For the extracted character code string of the financial institution name, the type name processing unit 25A of the matching processing unit 25 executes the matching processing algorithm using the type name of the financial institution as a predetermined number of candidates as in the case of the application form 20A. Then, a predetermined number of type names are specified (S9). This specifying process will be described with reference to FIGS. In the case of the application form 20B, the type is specified with reference to the correspondence table prepared in advance from the code data of the type.
[0037]
Next, in the case of the application form 20A, the unique name processing means 25B uses the recognized financial institution name character code string other than the identified financial institution type name as the unique name of the financial institution, The unique name corresponding to the identified type name of the financial institution is identified with reference to the financial institution unique name table as a predetermined number of candidates to execute a matching processing algorithm (S10). This unique name specifying process will be described with reference to FIGS. Further, in the case of the application form 20B, a matching process for identifying a specific name corresponding to the type name of the financial institution specified by the code data of the above type as a predetermined number of candidates with reference to the specific name table of the financial institution. Run the algorithm to identify the unique name. Then, based on the identified type name and unique name of the financial institution, the code extracting means 16 refers to the created financial institution unique name table and extracts a predetermined number of financial institution codes (S11).
[0038]
Subsequently, in FIG. 7, the display processing unit 27 executes a process for displaying the specified financial institution name as shown in FIG. 5 on the display unit 14 while indicating a predetermined number of candidates (S12). In the code search result screen as shown in FIG. 5, the name of the financial institution whose input can be changed by clicking the “OK” button according to the confirmation by the operator or the change of the financial institution name and the financial institution code as appropriate. The financial institution code with the name of the financial institution displayed in the area of the financial institution code is stored as a fixed code in, for example, the temporary buffer of the output file creating means 18 (S13).
[0039]
Based on the extraction setting, a character code string corresponding to the store name is extracted from the corresponding target character code data (S14), and the type name processing means of the matching processing means 25 is applied to the character code string of the store name. Assuming that 25A is the application form 20A, the matching process algorithm for specifying the store type name (FIG. 3B) as a predetermined number of candidates is executed to specify the store type name (S15). In the case of the application form 20B, the type of the store is specified with reference to the correspondence table prepared in advance from the code data of the type of the store.
[0040]
Further, the unique name processing means 25B corresponds to the type of the store specified as the unique name of the store in the character code string of the store name as a unique name of the store as in the case of the application form 20A. Refer to the store unique name table for the unique name or the unique name corresponding to the store type specified by the store type code data above, and execute the matching algorithm that identifies as a predetermined number of candidates and execute the unique name Is specified (S16). Further, in the case of the application form 20B, a matching processing algorithm that identifies a unique name corresponding to the store name of the financial institution specified by the code data of the store as a predetermined number of candidates with reference to the store unique name table To identify the unique name.
[0041]
Based on the identified store name and financial institution name, the code extracting means 26 refers to the created store unique name table and extracts a store code (S17). Moreover, the process for displaying the store name specified by the display processing means 27 on the display means 14 with a predetermined number of candidates is executed (S18). This code search result screen is also displayed in the store name and store code area where the input can be changed by clicking on the “OK” button according to the confirmation by the operator or appropriately changing the store name and store code. For example, the store male code of the store name is stored in the temporary buffer of the output file creation means 18 as a confirmation code (S19).
[0042]
The above processing is performed for all the recognized character code data to be extracted based on the extraction setting. When there are next processing candidates, the above steps S8 to S19 are repeated (S20), When the code search is performed on the processing candidates, the output file creation means 18 creates the extracted financial institution code and store code files for all the recognized character code data of the object and outputs it as the output file 16. (S21).
[0043]
Next, FIG. 8 shows a schematic explanatory diagram of the entire matching process according to the present invention. FIG. 8 is for the application form 20A shown in FIG. In FIG. 8, first, the character string of the recognized character code data of the input financial institution name is, for example, “A”, and the number of characters of the character string A is “a” (S31). Therefore, the type name of the character string A is specified, and a predetermined number of unique name parts (parts other than the type name) corresponding to “A” are taken out (S32). For example, a predetermined number of unique name portions B1, B2,... When the type names (M1, M2,...) Are specified are extracted. Then, a predetermined number of unique names indicated by the unique name portion of the character string A are specified (S33). For example, a predetermined number of unique names P11, P12... When the type name (M1) is specified are extracted, and a predetermined number of specific names P21, P22. A predetermined number of unique names are extracted according to the number of type names specified below. The identification of the type name and unique name will be described with reference to the following flowchart.
[0044]
9 to 12 show flowcharts of processing for specifying the financial institution type name. FIG. 9 is a schematic flowchart of the entire process for identifying a financial institution type name. In the type name processing unit 25A of the matching unit 25, first, the type of the recognized character code data of the input financial institution name is classified. A determination process is performed for the replacement name (S41), and it is determined whether or not they are completely matched (S42). As a result of the determination, the process is terminated when there is a complete match. Details of this processing will be described with reference to FIG.
[0045]
If it is not an exact match, a determination process is performed for the mixed name of the type (S43), and details of this process will be described with reference to FIG. Subsequently, a determination process is performed for a missing type name (S44), and details of this process will be described with reference to FIG. If the type can be determined in the determination process for the mixed name and the missing name (S45), the process ends. If the type cannot be determined, the entire name of the financial institution is set as “no type name”. (S46).
[0046]
Therefore, in the determination of the type of replacement name in FIG. 10, first, when the determination is made for all the type names (S51), the process ends. If the type replacement name has not been determined, the type name of the financial institution to be determined is T, for example, for the character code string (A: number of characters a) of the financial institution name in the input recognized character code data, and the number of characters Is m (S52). For example, when the type name T is “credit union”, the number of characters m is “4”.
[0047]
Therefore, the attribute of the type name T is determined from the association included in the program shown in FIG. 3A (S53), and in the case of the head, the number of characters a of the input character code string (A) is determined by the financial institution. It is determined whether the number of characters of the type name T is greater than m (S54). When the character number a is smaller than the character number m of the type name T, the process returns to S51 at the start of processing. If the character number a is larger than the character number m of the type name T, the number of characters at the beginning m of the character code string (A) is extracted, and the character match number e compared with the type name T is obtained (S55). For example, when the type name T to be matched is “trust union”, {m = 4} characters are extracted from the head of the input character code string (A), and “trust”, “for”, “set” The matching number e is obtained by performing a matching process to determine whether the character codes “” and “go” are included.
[0048]
Subsequently, it is determined whether or not the number of matches e is “0” (S56). If e = 0, the process returns to S51 at the start of processing. When the number of matches e is not “0”, the type name T is determined as the type name of the character code string A, and the unique name at this time is the character at the end {a−m} of the character code string A. The coincidence rate is represented by {e / m} (S57). Subsequently, it is determined whether the number of matches e is equal to the number of characters m (S58). If not equal, the process returns to S51 at the start of processing. If they are equal, {e = m} is determined as perfect match (match rate 1.0) (S59).
[0049]
On the other hand, in S53, even in the last case, it is determined whether the number of characters a in the input character code string (A) is larger than the number m of characters in the type name T of the financial institution (S60). When the character number a is smaller than the character number m of the type name T, the process returns to S51 at the start of processing. If the character number a is larger than the character number m of the type name T, the end m of the character code string (A) is taken out, and the number of matching characters e compared with the type name T is obtained (S61).
[0050]
Subsequently, it is determined whether or not the number of matches e is “0” (S62). If e = 0, the process returns to S51 at the start of processing. If the number of matches e is not “0”, the type name T is determined as the type name of the character code string A, and the unique name at this time is the character at the beginning of the character code string A {am}. The coincidence rate is represented by {e / m} (S63). Subsequently, it is determined whether the number of matches e is equal to the number of characters m (S64). If not equal, the process returns to S51 at the start of processing, and if equal, {e = m} is determined to be a perfect match (match rate 1.0). (S65).
[0051]
Next, in the mixed name type determination of FIG. 11, when all the type names are determined (S71), the process is terminated. If the mixed name of the type has not been determined, the type name of the financial institution to be determined is, for example, T for the character code string (A: number of characters a) of the financial institution name in the input recognized character code data, and the number of characters Is m (S72).
[0052]
Therefore, the attribute of the type name T is determined from the association included in the program shown in FIG. 3A (S73), and in the case of the top, the number of characters a of the input character code string (A) is determined by the financial institution. It is determined whether there is more than the number of characters {m + 1} obtained by adding one character to the number m of characters of the type name T (S74). If the number of characters a is smaller than the number of characters {m + 1}, the process returns to S71 at the start of processing. If the number of characters a is larger than the number of characters {m + 1}, the number of characters at the beginning {m + 1} of the character code string (A) is extracted, and the number of matching characters e compared with the type name T is obtained (S75).
[0053]
At this time, since there are a plurality of mixed names, the one with the largest number of matches is adopted. For example, when the type name T to be matched is “trust union”, {m + 1 = 5} characters are extracted from the head of the input character code string (A), and “trust”, “for”, “set” The matching number e is obtained by performing a matching process to determine whether the character codes “” and “go” are included.
[0054]
Subsequently, it is determined whether or not the number of matches e is “0” (S76). If e = 0, the process returns to S71 at the start of processing. When the number of matches e is not “0”, the type name T is determined as the type name of the character code string A, and the unique name at this time is the end {a− (m−1) of the character code string A. )} And the matching rate is represented by {e / m} (S77).
[0055]
On the other hand, in S73, even in the last case, it is determined whether the number of characters a in the input character code string (A) is larger than the number of characters {m + 1} obtained by adding one character to the number m of characters in the type name T of the financial institution (S78). ). If the number of characters a is smaller than the number of characters {m + 1}, the process returns to S71 at the start of processing. When the character number a is larger than the character number {m + 1}, the end {m + 1} of the character code string (A) is extracted, and the character match number e compared with the type name T is obtained (S79). Even in this case, since there are a plurality of mixed names, the one with the maximum number of matches is adopted.
[0056]
Subsequently, it is determined whether or not the number of matches e is “0” (S80). If e = 0, the process returns to S71 at the start of processing. If the number of matches e is not “0”, the type name T is determined as the type name of the character code string A, and the unique name at this time is the head of the character code string A {a− (m−1). )} And the matching rate is represented by {e / m} (S81).
[0057]
Next, in the determination of the type of missing names in FIG. 12, first, when all types of names are determined (S91), the process ends. If the determination of the missing name of the type has not been made, the type name of the financial institution to be determined is, for example, T for the character code string (A: number of characters a) of the financial institution name in the input recognized character code data, and the number of characters Is m (S92).
[0058]
Therefore, the attribute of the type name T is determined from the association included in the program shown in FIG. 3A (S93), and in the case of the head, the number of characters a of the input character code string (A) is determined by the financial institution. It is determined whether there is more than the number of characters {m−1} obtained by subtracting one character from the number m of characters of the type name T (S94). If the number of characters a is smaller than the number of characters {m−1}, the process returns to S91 at the start of processing. If the number of characters a is larger than the number of characters {m−1}, the number of characters {m−1} at the beginning of the character code string (A) is extracted, and the number of matching characters e compared with the type name T is obtained (S95). ). In this case, since there are a plurality of missing names, the one with the maximum number of matches is adopted.
[0059]
Subsequently, it is determined whether or not the number of matches e is “0” (S96). If e = 0, the process returns to S91 at the start of processing. When the number of matches e is not “0”, the type name T is determined as the type name of the character code string A, and the unique name at this time is the end {a− (m + 1)} of the character code string A. The match rate is represented by {e / m} (S97).
[0060]
On the other hand, in S93, even in the last case, it is determined whether the number of characters a in the input character code string (A) is larger than the number of characters {m-1} obtained by subtracting one character from the number m of characters in the type name T of the financial institution. (S98). If the number of characters a is smaller than the number of characters {m−1}, the process returns to S91 at the start of processing. If the number of characters a is larger than the number of characters {m−1}, the number of characters at the end {m−1} of the character code string (A) is extracted, and the number of matching characters e compared with the type name T is obtained (S99). ). Also in this case, since there are a plurality of missing names, the one with the largest number of matches is adopted.
[0061]
Subsequently, it is determined whether or not the number of matches e is “0” (S100). If e = 0, the process returns to S91 at the start of processing. If the number of matches e is not “0”, the type name T is determined as the type name of the character code string A, and the unique name at this time is the head of the character code string A {a− (m + 1)}. The matching rate is represented by {e / m} (S101).
[0062]
In this way, for the input recognition character code data of the predetermined financial institution, the type name is determined by determining the type replacement name and appropriately determining the mixed name and missing name. The result is determined in S45 in FIG. Then, the part other than the specified type name is set as the unique name, and when the type name is not specified, “no type name” is set and the matching process of FIGS. 13 to 18 by the specific name processing unit 25B in the matching processing unit 25 is performed. Transition.
[0063]
Next, FIGS. 13 to 18 show flowcharts of processing for specifying a unique name of a financial institution name. FIG. 13 is a schematic flowchart of the entire process for identifying the unique name of the financial institution. The unique name processing means 25B of the matching means 25 is identified as described above for the input recognized character code data of the financial institution name. For the unique name portion other than the type name, the unique name is judged to be completely matched with the financial institution unique name in the later-described financial institution unique name table (S111). As a result of the determination, if there is a completely unique name (S112), the process ends. Details of this processing will be described with reference to FIG. If there is no exact matching unique name (S112), a determination process is performed when the unique name is not perfect matching (S113). Details of this process will be described with reference to FIGS.
[0064]
Therefore, in FIG. 14, first, when the determination process is performed on all the pairs of the type name and the unique name obtained by the type name matching process S32 shown in FIG. 8 (S121), The process ends. Here, in the type name matching process, among the specified number of type names specified above, M is the type name of the target financial institution name, and B is the unique name portion of the character code string A corresponding thereto. (S122). Then, it is referred to whether or not there is a unique name P that matches B as the unique name in the type name M of the created financial institution unique name table (S123).
[0065]
As a result of the reference (S124), if it does not exist, the process returns to S121 of the process start. If it exists, the unique name P is determined as the input unique name, and since it is a perfect match, the match rate is “1.0. (S125). At this time, if the match rate at the time of specifying the type is “1.0”, it is determined that the match is complete (S127), and the process returns to S121 of the process start. If the type-specific matching rate is not “1.0”, the process returns to S121 of the process start.
[0066]
Therefore, in FIG. 15, first, when the determination process is performed for all the pairs of the type name and the unique name obtained by the type name matching process S32 shown in FIG. 8 (S131), The process ends. Among the specified number of type names specified above, the type name of the target financial institution name is M, the corresponding unique name portion of the character code string A is B, and the number of characters of B is b. (S132).
[0067]
Then, a matching process is performed to determine whether or not there is a unique name P that matches B as the unique name in the type name M of the created financial institution unique name table. After matching processing is performed for all the unique names of the type name M of the table, the process returns to S131 of the process start (S133). Therefore, when matching processing is performed for the unique name of the classification name M in the financial institution unique name table, the unique name of the classification name M in the financial institution unique name table is set to P and the number of characters is set to n (S134).
[0068]
Subsequently, a determination process is performed by matching the unique name B with the unique name P in the financial institution unique name table (S135). Details of this processing will be described with reference to FIG. Also, a determination process is performed by matching the mixed name of B as the unique name with the unique name P in the financial institution unique name table (S136). Details of this processing will be described with reference to FIG. Further, a determination process is performed by matching the missing name of B as the unique name with the unique name P in the financial institution unique name table (S137). Details of this processing will be described with reference to FIG. And these S134-S137 perform the determination process by matching about all the specific names P of the classification name M of the said financial institution specific name table.
[0069]
Therefore, in FIG. 16, as a matching process corresponding to S135 in FIG. 15, first, the number b of characters B as the unique name portion of the character code string A and the unique name P in the type name M of the financial institution unique name table are listed. (S141), if the character number b of B is not the same as the character number n of the unique name P, the process is terminated, and if it is the same, the matching number e of the comparison result is obtained ( S142). Then, it is determined whether or not the number of matches e is “0” (S143). If “0”, the process returns to S141. If it is not “0”, the unique name P is determined as the unique name of the character code string A, The match rate (e / n) is obtained and the process ends (S144).
[0070]
In FIG. 17, the number of characters {b + 1} obtained by adding one character to the number of characters b of B as the unique name portion of the character code string A and the unique name P in the type name M of the financial institution unique name table. The number n of characters is compared (S151). If the number of characters {b + 1} is not the same as the number of characters n of the unique name P, the process is terminated. If the number is the same, the number of matches e as a result of the comparison is obtained (S152). ). In this case, since there are a plurality of mixed names, the maximum matching number e is adopted. Then, it is determined whether the number of matches e is “0” (S153). If “0”, the process returns to S151. If it is not “0”, the unique name P is determined to be the unique name of the character code string A, The match rate (e / n) is obtained and the process ends (S154).
[0071]
Subsequently, in FIG. 18, the number of characters {b−1} obtained by subtracting one character from the number of characters b of the character code string A as the unique name portion of B and the unique name in the type name M of the financial institution unique name table. If the number of characters {b-1} is not the same as the number of characters n of the unique name P, the process is terminated. If the number is the same, the number of matches e as a result of the comparison is compared. Is obtained (S162). In this case, since there are a plurality of missing names, the maximum matching number e is adopted. Then, it is determined whether or not the number of matches e is “0” (S163). If “0”, the process returns to S161. If it is not “0”, the unique name P is determined as the unique name of the character code string A, The match rate (e / n) is obtained and the process ends (S164).
[0072]
By the way, in FIG. 9, in the classification determination, the entire character code string of the input financial institution is determined as a unique name for the determination result of “no classification name”, and the unique name (for example, D, number of characters) is determined. And FIG. 13 to FIG. 18 show the determination processing by matching whether there is a unique name P that matches D as the proper name among the type names in the created financial institution proper name table. (However, S133 in FIG. 15 is processed for all types).
[0073]
As described above, the code extraction unit 16 refers to the created financial institution unique name table based on the type name and unique name of the financial institution identified in the matching process of FIGS. 9 to 18 and the unique name without the type name. Then, a predetermined number of financial institution codes are extracted, and the display processing means 27 performs processing to display a screen as shown in FIG. As a result, the financial institution code assigned to the financial institution can be extracted from the financial institution name of the entry as shown in FIG. 2, and the efficiency of the paperwork when the individual code is given can be improved. Is.
[0074]
In addition, the store name matching process specifies a predetermined number of store names (or no store names) using the store unique name table created above for the store names corresponding to the specified financial institution names, and corresponds to the store names. By identifying the unique name, the store code of the store can be extracted. This store name matching process is basically the same as the above-described financial institution name matching process, and a description thereof will be omitted.
[0075]
Next, FIG. 19 shows an explanatory diagram of another embodiment according to the present invention. FIG. 19A shows a screen for displaying the predetermined number of financial institution names (store names) and corresponding financial institution codes (store codes) on the screen of the code search result shown in FIG. An “image” button that is a display button for displaying image data of a financial institution name is provided. In other words, the image data of the target financial institution name is provided as the image data 19A of the recognition data 19 by the data recognition means 17, and the data extraction means 24 is described above when the “image” button is clicked. The image data of the file associated with the ID number is extracted from the image data 19A, and the display processing unit 27 causes the window on which the image is displayed to appear on the screen.
[0076]
Therefore, in FIG. 19B, first, when the financial institution name specified by the display processing means 27 is displayed in the “financial institution” area by naming predetermined candidates according to the matching rate as described above ( In S171), when an “image” button is clicked and an image signal is input (S172), the data extraction means 24 uses, for example, an ID number in the corresponding request form (application form 20) from the image data 19A in the recognition data 19. The image data of the associated file is extracted and extracted, and displayed in a separate window, for example (S173).
[0077]
If the “image” button is not clicked in step S172, or if the operator confirms the displayed image and inputs an appropriate change in the “changeable” area of the “financial institution” area on the screen. When a confirmation signal is input by clicking the “OK” button after confirmation and change by the operator, the displayed financial institution code is associated with the name of the financial institution and stored as a confirmation code (S174).
[0078]
As described above, by providing the “image” button on the display screen, it is easy for the operator to confirm the financial institution code (store code) of the financial institution name (store name), and a reliable individual code can be assigned. Is.
[0079]
【The invention's effect】
  As described above, claims 1 to3According to the invention, for the recognized character code data of the organization name including the name belonging to the lower order entered in the predetermined form,Predetermined for either the beginning or end of the recognized character code dataMatching with the category name (character code) is performed to identify the category name, and the unique name corresponding to the identified category name is matched with the character code of the unique name obtained from the organization code table for the unique name. Identifying a predetermined number of group names according to the degree of matching of the specified type name and unique name, and extracting a group code including a predetermined number of group sub-codes of the specified number of group names from the table Instead of displaying the data and outputting a data file with the organization code determined as appropriate to the organization name, the individual code assigned to the organization can be extracted from the organization name of the entry. It is possible to improve the efficiency of the paperwork in the case of granting.
[0080]
  Claim4According to the invention, the display processing means is provided with a display button for displaying the image data of the target organization name on the screen for displaying the specified organization name and organization code, thereby facilitating confirmation by the operator and reliable individual code. It can be given.
[Brief description of the drawings]
FIG. 1 is a block configuration diagram of a code search processing system according to an embodiment of the present invention.
FIG. 2 is an explanatory diagram of a form book that is a target of code search processing;
FIG. 3 is an explanatory diagram showing an example of types of recognition type names in code search processing.
4 is an explanatory diagram showing recognition patterns of type names and unique names in FIG. 3. FIG.
FIG. 5 is an explanatory diagram showing a screen display of an example of a code search result.
FIG. 6 is a flowchart (1) of the entire code search process according to the present invention.
FIG. 7 is a flowchart (2) of the entire code search process according to the present invention.
FIG. 8 is a schematic explanatory diagram of an entire matching process according to the present invention.
FIG. 9 is a flowchart (1) of processing for identifying a financial institution type name.
FIG. 10 is a flowchart (2) of processing for identifying a financial institution type name.
FIG. 11 is a flowchart (3) of processing for identifying a financial institution type name.
FIG. 12 is a flowchart (4) of processing for identifying a financial institution type name.
FIG. 13 is a flowchart (1) of processing for identifying a unique name of a financial institution name.
FIG. 14 is a flowchart (2) of a process for specifying a unique name of a financial institution name.
FIG. 15 is a flowchart (3) of a process for specifying a unique name of a financial institution name.
FIG. 16 is a flowchart (4) of a process for specifying a unique name of a financial institution name.
FIG. 17 is a flowchart (5) of a process for specifying a unique name of a financial institution name.
FIG. 18 is a flowchart (6) of processing for specifying a unique name of a financial institution name.
FIG. 19 is an explanatory diagram of another embodiment according to the present invention.
[Explanation of symbols]
11 Code search processing system
12 computers
13 Input means
14 Display means
15 Financial Institution Information DB
16 output files
19 Recognition data
20 Application form
23 Table creation means
24 Data extraction means
25 Matching processing means
25A Type name processing means
25B proper name processing means
26 Code extraction means
27 Display processing means
28 Output file creation means

Claims (4)

Based on the recognized character code data, which is at least a predetermined organization name entered in a predetermined form, and which is read as an image data of an organization name including a predetermined number of lower names belonging to the organization, and recognized as a character code string A code search processing system that recognizes the name of the organization name and the unique name of the organization name excluding the category, and obtains the organization name code including the predetermined number of organization subcodes assigned to the organization name from the organization code table Because
The recognition character code data indicating the target group name is subjected to a matching process with a predetermined number of character codes of a predetermined number of names at the beginning or end of the recognition character code data , and a predetermined number of type names are obtained. Identify the unique name by sequentially matching with the character code of the unique name obtained with the specified category name by referring to the organization code table with the remaining part of the identified category name as the unique name. Matching processing means for performing a process of specifying a predetermined number of group names according to the degree of matching of the entire group name of the identified type name and unique name,
Code extracting means for extracting a predetermined number of corresponding group codes from the group code table based on the specified number of group names specified;
Display processing means that can be changed by external input, and performs processing for displaying the specified predetermined number of organization names and corresponding organization codes;
A code search processing system comprising: A claim 1 Symbol placement code retrieval processing system, the organization code table, the information in a given predetermined groups, and collective name given number of about said predetermined groups, a predetermined number of groups corresponding thereto A code search processing system characterized by being created in association with a code. 3. The code search processing system according to claim 1 or 2 , wherein a data file of a confirmed organization name and organization code is generated and output in response to a confirmation signal from outside for the display in the display processing means. A code retrieval processing system comprising output file creation means for performing A claim 1-3 least one code search processing system according to the display processing unit, a screen for displaying the organization name and the corresponding organization code of a predetermined number of identified image of the subject Organization A code search processing system comprising a display button for displaying data.

Images