JP6664201B2 - Abutting processing device, abutting processing method and abutting processing program - Google Patents

Description

  The present invention relates to a matching processing device, a matching processing method, and a matching processing program for performing a matching process on a plurality of databases each storing a plurality of records.

  Conventionally, when a plurality of records included in each of a plurality of databases (for example, a database A and a database B) are matched, a matching result is obtained by assigning a code for classifying the matching state to each record. It should be noted that a record in the present specification is a group of data stored in a database, and indicates a group of data indicating unique information such as, for example, address information. In addition, the term “matching” as used in the present specification means that databases are compared with each other (or between records of a database) in order to check whether a part or all of data included in each record of a plurality of databases match. Is to match.

  In the conventional matching process, for example, when a record in the database A matches a record in the database B or when some association is obtained, a code “1” indicating that the record matches the record is used. , And a code "-1" indicating that the record is a non-coincidence is added when the records do not coincide with each other.

  In addition, as conventional techniques, there are techniques disclosed in Patent Documents 1 to 3 below. Patent Literature 1 discloses a technique in which a plurality of databases storing character string hierarchical data (such as an address name) and code character string data (such as a postal code) are combined to create a combined database having a tree structure. ing. According to the technology disclosed in Patent Literature 1, character string hierarchical data is data having a hierarchical structure such as an address name, and a match / mismatch is determined by comparing the character string hierarchical data in a layer unit (node unit).

  Further, Patent Literature 2 discloses a technique of merging (integrating) an address code database (including address information represented by an array of numbers) with an address database (including place name information) to create a merge database. It has been disclosed. According to the technology disclosed in Patent Literature 2, when merging the databases, the address (the name of the town and village) is deleted one character at a time from the right end and searched, and it is determined that the character string of the hit location is coded. And memorize. At this time, for example, a symbol “/” is inserted to identify a coded portion.

  Further, Patent Literature 3 discloses a technique for searching for address data in an address database based on a search address character string input by a user. According to the technology disclosed in Patent Literature 3, if there is no address corresponding to the search address character string input by the user, the search address character string is rearranged by a predetermined amount (for example, until one or more addresses are hit). Address) and repeat the search.

JP 2010-134828 A (abstract, paragraph [0116], FIG. 1, FIG. 30, FIG. 31, FIG. 34) JP-A-2003-167912 (abstract, paragraph [0028], FIG. 6) JP 2003-186880 A (abstract)

  In the related art, identification of a matching / mismatching state in each record (for example, the above-mentioned code “1” or “−1” is added), and identification and correction of a portion to be corrected for each mismatched record (for example, (Disclosed technologies of Patent Documents 1 and 2 described above). However, even if a part to be corrected is specified for each record, it is not easy to find an appropriate correction value. Also, even if a mismatched record is identified, it is necessary to know where the cause of the mismatch and how many mismatched records occur in the entire record in the database for the same reason. It is difficult to efficiently and appropriately correct the mismatched records.

  Further, in the disclosed techniques of Patent Literatures 1 to 3, a forward search method for searching for character string information such as address information from the front or a backward search method for searching from behind is mainly used. However, such a forward search method or a backward search method uses, for example, when a plurality of differences (a plurality of unmatched portions) exist between records of a database to be compared, a plurality of locations where these differences occur. May not be detected efficiently and accurately.

  In view of the above problems, the present invention matches a plurality of records included in each of a plurality of databases, and outputs a matching result including a portion where a match / mismatch occurs in each record. It is an object of the present invention to provide a matching processing device, a matching processing method, and a matching processing program that can clearly indicate the state of match / mismatch between records in databases.

In order to achieve the above object, a matching processing device of the present invention is a matching processing device that performs a matching process on a plurality of databases each storing a plurality of records including character string information,
A segment setting unit that sets a plurality of segments according to a predetermined rule for the character string information;
The character string information included in the record in the first database of the plurality of databases is compared with the character string information included in the record in the second database of the plurality of databases. A record comparing unit that detects a match or mismatch of character string information in a segment;
When a record including the common character string information in all of the compared segments is present in both the first and second databases, information indicating that the common character string information is included in all of the compared segments is stored. A matching result generation unit for associating with each of the records in the first and second databases including character string information common to all of the compared segments;
Have.

In order to achieve the above object, a matching processing method of the present invention is a matching processing device that performs a matching process on a plurality of databases each storing a plurality of records including character string information,
A segment setting step of setting a plurality of segments according to a predetermined rule for the character string information;
The character string information included in the record in the first database of the plurality of databases is compared with the character string information included in the record in the second database of the plurality of databases. A record comparison step of detecting a match or mismatch of character string information in a segment;
When a record including the common character string information in all of the compared segments is present in both the first and second databases, information indicating that the common character string information is included in all of the compared segments is stored. Generating a matching result associated with each of the records in the first and second databases including character string information common to all of the compared segments.
Have.

Further, in order to achieve the above object, a match processing program according to the present invention includes a match processing method for causing a computer to execute a match processing method for performing a match processing of a plurality of databases each storing a plurality of records including character string information. A processing program,
A segment setting step of setting a plurality of segments according to a predetermined rule for the character string information;
The character string information included in the record in the first database of the plurality of databases is compared with the character string information included in the record in the second database of the plurality of databases. A record comparison step of detecting a match or mismatch of character string information in a segment;
When a record including the common character string information in all of the compared segments is present in both the first and second databases, information indicating that the common character string information is included in all of the compared segments is stored. Generating a matching result associated with each of the records in the first and second databases including character string information common to all of the compared segments.
This is a program for causing a computer to execute the abutting method.

  The present invention has the above-described configuration or processing, matches a plurality of records included in each of a plurality of databases, and outputs a matching result including a portion where a match / mismatch occurs in a mutual record. This has the effect that the status of the match / mismatch between the records in the databases can be clearly shown. Further, the present invention has an effect that it is possible to confirm the details of the status of the match / mismatch by quantitatively indicating the status of the match / mismatch of each record between the databases. Further, the present invention has an effect that the work for resolving the inconsistency between databases can be promptly performed for the inconsistent records.

  Further, the present invention has a remarkable effect particularly when applied to a reconciliation process between a ledger database including address information and a map database including address information associated with image information representing a map or the like. In the case of multiple databases containing address information, in particular, there is the problem of inconsistencies in the lot numbers, etc. in the national land. Had become. The present invention is to quickly find an appropriate key item value by using a code system that clearly indicates a match / mismatch situation as described in the embodiment of the present invention in the examination of a land register and an address. And it is possible to know the cause of the mismatch.

It is a block diagram showing an example of composition of a matching processing device common to an embodiment of the invention. It is a flowchart which shows the outline | summary of the matching process common to embodiment of this invention. It is a flowchart which shows an example of the pre-setting process in the matching process common to the embodiment of the present invention. It is a flowchart which shows an example of the joining code setting process in the joining process common to the embodiment of the present invention. It is a figure showing an example of databases A and B used for explaining matching processing in an embodiment of the invention. It is a figure showing an example of the 1st state of databases A and B in the processing process of the matching processing in an embodiment of the invention. FIG. 5 is a diagram schematically illustrating an example of a filter when three segments are set for character string information, and an example of a matching code set corresponding to the filter in the embodiment of the present invention. . FIG. 5 is a diagram schematically illustrating an example of a filter when four segments are set for character string information, and an example of a matching code set corresponding to the filter in the embodiment of the present invention. . FIG. 7 is a diagram schematically illustrating an example of a filter when five segments are set for character string information, and an example of a matching code set corresponding to the filter in the embodiment of the present invention. . It is a figure showing an example of the 2nd state of databases A and B in the processing process of matching processing in an embodiment of the invention. It is a figure showing an example of the 3rd state of databases A and B in the processing process of the matching processing in an embodiment of the invention. FIG. 7 is a diagram illustrating an example of a state of databases A and B including a final matching result obtained in a matching process according to the embodiment of the present invention. FIG. 9 is a diagram illustrating an example of a case where the match code is further extended to have information indicating the number of records in the state illustrated in FIG. 8. FIG. 13 is a diagram illustrating an example of a state where only records in which a specific match code “11_” is set are extracted from the final match result in the embodiment of the present invention. FIG. 13 is a diagram showing an example of a state where only records in which a specific match code “11_11” is set are extracted from the final match result in the example according to the embodiment of the present invention. FIG. 13 is a diagram showing an example of a state where only records in which a specific match code “11_1_” is set are extracted from the final match result in the example according to the embodiment of the present invention. FIG. 13 is a diagram illustrating an example of a state where only records in which a specific match code “111_1” is set are extracted from the final match result in the example according to the embodiment of the present invention. FIG. 12 is a diagram illustrating an example of a state in which address information related to a fourth record in the table of FIG. 11 is displayed on a display screen by superimposing the address information on a map based on a map database. In the example according to the embodiment of the present invention, a state in which only a record in which a specific match code “11111: 11” is set is extracted from a final match result, and a state in which an “area” field is displayed It is a figure showing an example of. Regarding the records determined to correspond one-to-one in the two databases (ledger database and map database) by the reconciliation process according to the present invention, a common item (““ FIG. 21 is a diagram illustrating an example of a scatter diagram created using values of an “area” field).

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

  FIG. 1 is a block diagram showing an example of a configuration of a matching processing apparatus common to the embodiments of the present invention. The collision processing apparatus 10 illustrated in FIG. 1 includes a database reading unit 11 and a collision processing unit 12. Further, a storage medium 20 storing a plurality of databases, an operation input unit 30, and a display unit 40 are connected to the reconciliation processing apparatus 10.

  The database reading unit 11 has a function of reading a plurality of databases (databases A and B in FIG. 1) stored in the storage medium 20. The database read by the abutting processing device 10 is temporarily stored in, for example, a main storage device (main memory, not shown in FIG. 1) in the abutting processing device 10 so that the abutting processing unit 12 performs processing. You.

  The matching processing unit 12 has a function of performing a matching process on each record in a plurality of databases (databases A and B in FIG. 1) read by the database reading unit 11. And a matching result output unit 15.

  The record comparing unit 13 performs a matching process of comparing character string information included in each record of each of a plurality of databases (databases A and B in FIG. 1) with each other and detecting a match or mismatch of the character string information. Has a function. The record comparing unit 13 has a segment setting unit 16. The segment setting unit 16 has a function of segmenting the character string information included in each record according to a rule regarding a predetermined segment structure, and setting a plurality of segments for the character string information. The segmentation of the character string information is to divide the character string information into a plurality of parts (each part includes at least one character). In this specification, the divided parts are referred to as segments. Call. By segmenting the character string information, a plurality of segments are set for the character string information, for example, as shown in FIG. 4 described later.

  The matching result generating unit 14 has a function of generating a matching result for each record of a plurality of databases (databases A and B in FIG. 1) based on the comparison result by the record comparing unit 13. The matching result generation unit 14 can generate a matching result of each record of a plurality of databases (databases A and B in FIG. 1) as, for example, a matching code (code information integrating the matching results) described later. is there. The comparison result output unit 15 has a function of outputting the comparison result generated by the comparison result generation unit 14, recording the result as data in a database, and displaying the data on the display unit 40.

  The storage medium 20 has a function of storing a database, and can be realized by, for example, a hard disk, an optical recording medium, a magnetic recording medium, a nonvolatile memory, or the like. Note that the storage medium 20 may be arranged at a location accessible by the matching apparatus 10 via a network or the like. Although FIG. 1 shows a state in which two databases (databases A and B) are stored in the same storage medium 20 as a plurality of databases, the plurality of databases are stored in different storage media 20 respectively. May be. Hereinafter, a description will be given mainly of a case where the matching process is performed in two databases. However, the matching process may be performed in three or more databases.

  The database stored in the storage medium 20 is an aggregate of data, and has a plurality of data including character string information. The data included in the database is stored in the database in record units based on, for example, a predetermined storage format (for example, a table format). The character string information included in each record is any type of information, and may include, for example, address information having a hierarchical structure such as municipalities, large letters, small letters, lot numbers, and branch numbers. Further, the number of records included in a plurality of databases may be the same or different.

  The operation input unit 30 has a function of enabling a user operation in the matching processing apparatus 10, and can be realized by an input interface such as a mouse or a keyboard. By using the operation input unit 30, the user can specify a database to be read by the amalgamation processing device 10, set in the amalgamation process, instruct a start of the amalgamation process, display an information on the display unit 40, and various other settings. And instructions can be input to the matching processing apparatus 10.

  In addition, the display unit 40 has a function of providing data stored in each of the plurality of databases and a match result generated by the match processing unit 12 as information that can be visually recognized by a user. Alternatively, it can be realized by a display or the like. Further, the operation input unit 30 and the display unit 40 may be realized by a touch panel or the like in which the operation input function and the display function are integrated.

  In FIG. 1, each function of the matching processing apparatus 10 is schematically illustrated by blocks, but each function represented by these blocks is, for example, hardware or a CPU (Central Processing Unit). (Device) executes software (program). The matching processing apparatus 10 can be realized by, for example, a general-purpose PC (Personal Computer), and each processing in the matching processing apparatus 10 is performed by each processing in the embodiment of the present invention (for example, see FIGS. 2A to 2C described later). The processing can be realized by the CPU executing a program in which an execution instruction of each process included in the illustrated flowchart is described.

  Although not shown in FIG. 1, the matching processing device 10 implemented by a general-purpose PC includes a main storage device (main memory) for temporarily storing data, programs, and the like. The processing in the embodiment of the present invention is performed while temporarily storing data in the database read by the unit 11, data changed or generated during processing, programs to be executed by the CPU, and the like in the main storage device. It is possible.

  Next, the matching process in the embodiment of the present invention will be described with reference to FIGS. 2A to 2C. FIG. 2A is a flowchart showing an outline of the abutting process common to the embodiment of the present invention, and FIG. 2B is a presetting process in the abutting process common to the embodiment of the present invention (the advance setting process of step S100 shown in FIG. 2A). FIG. 2C is a flowchart showing an example of a match code setting process (match code setting process of step S200 shown in FIG. 2A) in the match process common to the embodiment of the present invention. Each process of the flowcharts shown in FIGS. 2A to 2C is executed by the matching processing device 10 shown in FIG. In the following, in order to facilitate understanding of the present invention, a description will be given by taking as an example a case where each record of two databases A and B to be subjected to the matching process includes the character string information shown in FIG. Do

  As shown in FIG. 2A, the matching process in the embodiment of the present invention is roughly classified into a pre-setting process (step S100) and a matching code setting process (step S200). In the pre-setting processing of step S100, setting of parameters used in the subsequent matching code setting processing is performed. In the match code setting process in step S200, records in a plurality of databases (for example, databases A and B) that are actually subjected to the match process are compared, and information indicating a match result (for example, A process of giving a matching code) is performed.

  In the pre-setting process, as shown in FIG. 2B, the matching processing device 10 reads out the two databases A and B stored in the storage medium 20 and matches each record included in these databases A and B. A code column is initialized (for example, “0” is set as an initial value), a common segment structure is set for each record, and a filter according to the segment structure is set (step S101). The set segment structure and filter information are held by the matching processing apparatus 10 for use in subsequent processing.

  The segment structure set in step S101 is a structure in which character string information included in each record is segmented for each predetermined number of characters. Can be regarded as being composed of segments. For example, as shown in FIG. 4, for each record (including 10-digit character string information) of databases A and B, a segment including a character string of the upper 4 digits (the first to fourth digits from the left). S1, a segment S2 including a character string of middle three digits (fifth to seventh digits from the left) and a segment S3 including a character string of lower three digits (eighth to tenth digits from the left) Can be set.

  In step S101, an arbitrary segment structure may be set according to the length (number of characters) of the character string information included in each record of the databases A and B and the position of the character string information to be compared in the subsequent matching code setting processing. Can be set. For example, in the example of FIG. 4, three segments are set for the character string information, and the length of each segment is set to four digits, three digits, and three digits in order from the left. The number and length can be set arbitrarily. In the example of FIG. 4, the segments are set so as to cover the entire character string information (that is, the sum of the lengths of the respective segments is equal to the entire length of the character string information). A segment may be set only for a part of the information (a part excluding the end or the middle). Further, as the segment structure set in step S101, a segment structure preset for character string information may be used. For example, when address information is included as character string information, it is possible to set hierarchies such as municipalities, large letters, small letters, lot numbers, and branch numbers as segments.

  The filter set in step S101 is used to set which segment combination of the segments included in the set segment structure is to be compared in the subsequent matching code setting process. For example, when three segments are set for the character string information as in the example of FIG. 4, it is possible to set a combination of the three segments as a comparison target as a filter corresponding to the segment structure. is there.

  FIG. 5A schematically shows an example of a filter when three segments are set for character string information as in the example of FIG. 4, and an example of a match code set corresponding to the filter. FIG. The vertical array (row) in FIG. 5A represents identification information (LVL = 1 to 7) of each filter, and the horizontal array (column) in FIG. 5A is a segment to be compared when each filter is used. (S1 to S3).

  The filter schematically illustrated in FIG. 5A is used when a record search is performed in the subsequent matching code setting process (the process of step S207 in FIG. 2C described later). The segment described as “comparison” in each filter indicates that it is determined whether or not common character string information exists for each record of the databases A and B. On the other hand, hatched segments indicate that no determination is made regarding common character string information. Specifically, for example, a filter with LVL = 1 is described as “comparison” in all the segments S1 to S3. Therefore, when the filter of LVL = 1 is used, it is determined whether or not there is common character string information in all the segments S1 to S3 (in this example, the entire character string information of 10 digits). Also, for example, in the filter of LVL = 2, the segment S3 is hatched, and the segment S1 and S2 are described as “comparison”. Therefore, when the filter of LVL = 2 is used, it is determined whether or not there is common character string information in both of the segments S1 and S2, while no determination is made for the segment S3.

  FIG. 5A shows a match code set in a record when common character string information is detected using each filter. This matching code has the same number of digits as the number of segments, and each digit of the matching code corresponds to the position of each segment. For example, in the example shown in FIG. 5A, the character string information is divided into three segments, and the matching code is set to three digits corresponding to each segment.

  The matching code is a code unique to each filter. Each digit of the matching code corresponds to each of the segments S1 to S3, and corresponds to the segment for determining whether or not common character string information exists (the segment described as “Compare” in FIG. 5A). The value “1” is set in the digit to be set. In addition, “_” (under bar) is set in a digit corresponding to a segment (segment indicated by hatching in FIG. 5A) for which it is not determined whether there is common character string information.

  Also, as for the match code, as a result of determining whether common character string information exists using the corresponding filter, records having common character string information in all segments set in the filter are stored in the database. If it exists in both A and B, it is set for these records. For example, a record in which the match code "111" is set indicates that a record having common character string information in all of the segments S1 to S3 is determined to exist in both the databases A and B. ing. Further, for example, a record in which the match code “_11” is set indicates that a record having common character string information is determined to exist in both the databases A and B in the segments S2 and S3. ing. Therefore, by referring to the matching code set in a certain record, the record is a record common to both the databases A and B, and the common record has common character string information in which segment. Can be grasped.

FIGS. 5B and 5C schematically show an example of a filter when four segments and five segments are respectively set for character string information, and an example of a match code set when they match. FIG. The number of filters (i.e., corresponding to the number of LVLs) is 2 ^N -1 maximum when N is the number of segments in the segment structure (up to 7 when N = 3 and up to 7 when N = 4). 15 and a maximum of 31 when N = 5). However, not all filters need to be used. In the matching code setting process described later, the filter of LVL = 1 is used first, and then the filter of LVL = 2, the filter of LVL = 3,... Are sequentially used while incrementing the value of LVL. . Although it is possible to arbitrarily change which LVL is set to a filter having what kind of segment combination, a filter having a larger number of segments to be compared (ie, described as “comparison” in FIGS. 5A to 5C) It is desirable that the matching code setting process using a filter having more locations is performed preferentially.

  5A to 5C described above are schematically shown to explain the concept of a filter set according to the present invention, and those skilled in the art will recognize such a filter set as a substantial one. It is not necessary to prepare as appropriate data, and it can be realized by appropriately executing a process such as making an appropriate segment included in the character string information a comparison target or excluding it from the comparison target.

  Further, in the preliminary preparation processing, the matching processing device 10 sets a filter (initial filter) used first in the subsequent matching code setting processing as a variable LVL (step S103). In step S103, an arbitrary filter can be set. For example, it is desirable to set, as the initial filter, a filter of LVL = 1, which has the largest number of segments to be compared in the filter set of FIG. 5A. In addition to the setting of the initial filter, the user may be able to set a filter used in the matching code setting process, a final filter used in the last process, and the like. Note that the processing may be set so that the process using all filters (ie, the process of comparing all combinations of segments) is automatically performed without the user explicitly setting the final filter.

  After the setting of the parameters and the like is completed in the above-described pre-setting processing, in the matching code setting processing, as shown in FIG. Then, referring to the character string information of each record in the segment specified by the initial filter, the number of records including the same character string information is calculated, and the maximum value of the number of records is extracted (step S201). For example, when the filter of LVL = 1 in FIG. 5A (the filter including the segments S1 to S3) is set as the initial filter, in the example of the database A shown in FIG. 4, the same character string information is used in the segments S1 to S3. The record including the record No. 5 and 6 (the same number of records = 2), that is, the maximum number of records is 2. Similarly, in the example of the database B shown in FIG. 4, records including the same character string information in the segments S1 to S3 are record Nos. There are 4 to 6 records (the same number of records = 3), and the maximum number of records is 3. The maximum value of the number of records in the database A is set to the value vAMAX, and the maximum value of the number of records in the database B is set to the value vBMAX. These values vAMAX and vBMAX are used as the number of loops of a loop process described later. In the loop processing to be described later, the processing is appropriately performed even when the number of loops is the total number of records in the respective databases A and B. However, the value vAMAX which is the maximum value of the number of records is previously determined. By setting vBMAX and vBMAX as the number of loops, the processing time required for loop processing can be reduced.

  As described above, when the setting of the maximum value (the number of loop times vAMAX and vBMAX) of the number of records of the initial filter and the final filter and the databases A and B is completed, the matching processing apparatus 10 It is possible to perform a record matching process. For example, the matching processing apparatus 10 first sets the number of identical records (variable A1) in the database A to 1 (step S203), and sets the number of identical records (variable B1) in the database B to 1 (step S203). S205), records matching these conditions are extracted from each of the databases A and B, and a common record (that is, a record constituted by the same character string information) is searched (step S207). If a common record having common character string information is not detected in step S207, it is determined that there is no record for which a match code should be set (step S209), and the processes in steps S211 and S213 are performed. Not executed.

  Referring to the examples of the databases A and B shown in FIG. Records having the same number of records of 1 in the database B are records No. 1 to 4, 7, and 8. These are records 1 to 3, 7, and 8. By comparing these records, the reconciliation processing apparatus 10 records the record No. in the database A. 2 and the record No. in the database B. The second record is detected as a common record having the common character string information “1235792384” in the segments S1 to S3.

  Then, the matching processing apparatus 10 sets a matching code corresponding to the currently set filter (here, the initial filter) for these records common to the databases A and B (step S211). For example, as shown in FIG. 5A, it is desirable to use a matching code indicating the position of a segment to be compared (that is, the position of a segment where common character string information is detected), as shown in FIG. 5A. Can be expressed by various methods. The matching code thus obtained is written in the matching code column of the corresponding record (step S213). As a result, when the matching code shown in FIG. 2 and the record No. in the database B. A match code “111” is set for the record of No. 2.

  After terminating the search for the variable A1 = 1 and the variable B1 = 1, the matching processing apparatus 10 subsequently determines whether the variable B1 is equal to the value vBMAX (step S215), and the variable B1 has not reached the value vBMAX. In this case, the variable B1 is increased (incremented) by 1 (step S217), and the processing after step S207 is performed again. If it is determined in step S215 that the variable B1 has reached the value vBMAX, it is determined whether the variable A1 is equal to the value vAMAX (step S219). If the variable A1 has not reached the value vAMAX, , The variable A1 is incremented (step S221), and the processing after step S205 is performed again. As described above, by the processing of steps S203 to S221, the variable A1 in the range from 1 to vAMAX is set for the set filter (here, the filter of LVL = 1 set as the initial filter), In each combination with the variable B1 existing in the range from 1 to vBMAX, it is searched whether or not a common record exists between the database A and the database B. A match code is set in the match code field.

  By this series of processing, in the example of the databases A and B in FIG. 2 and the record No. in the database B. In addition to the setting of the match code in the record of the record No. 2, the record No. 5 and 6 and the record No. in the database B. A match code "111" is set for records 4 to 6 (obtained under the conditions of variable A1 = 2 and variable B1 = 3). As a result, for example, the databases A and B are in the state shown in FIG.

  When the variable A1 has reached vAMAX and the variable B1 has reached vBMAX, and the processing for all the same number of records has been completed, the matching processing apparatus 10 determines whether or not the filter identification information (variable LVL) is equal to the final filter. If the variable LVL has not reached the final filter (step S223), the filter is changed (for example, the variable LVL is incremented) (step S225), and the processing from step S201 is performed again. For example, when LVL = 1 is incremented to LVL = 2, in the example of the filter set shown in FIG. 5A, the processing of step S201 and subsequent steps is performed using the filter of LVL = 2 corresponding to the match code “11_”. Will be done again.

  In the processing after step S201 executed when the variable LVL is incremented to level 2 (variable LVL = 2), and in the processing after step S201 executed with the variable LVL raised, the matching code column The record in which the matching code is set is excluded from the search target in the search processing in step S207. That is, once a match code is set for a record in the course of processing, control is performed so that the match code of that record is not changed.

  By performing the setting of the match code while changing the filter in this way, the state in which the match code obtained using the filter of LVL = 1 is set (the state shown in databases A and B in FIG. 6) is further increased. When a filter of LVL = 2 is used, record No. 7 and the record No. in the database B. A match code "11_" is set for records 7 and 8 (obtained under the condition of variable LVL = 2, variable A1 = 1, variable B1 = 2), and when a filter of LVL = 3 is used, A record No. in A 4 and the record No. in the database B. A match code "1_1" is set for record 3 (obtained under the condition of variable LVL = 3, variable A1 = 1, variable B1 = 1). When a filter of LVL = 5 is used, Record No. 1 and the record No. in the database B. A match code "1__" is set for one record (obtained under the condition of variable LVL = 5, variable A1 = 1, variable B1 = 1). When a filter of LVL = 6 is used, Record No. 3 and the record No. in the database B. A match code “_1_” is set for the record of No. 9 (obtained under the conditions of variable LVL = 6, variable A1 = 1, and variable B1 = 1). As a result, for example, the databases A and B are in the state shown in FIG.

  When the matching process from the initial filter (for example, the filter of LVL = 1 in FIG. 5A) to the final filter (for example, the filter of level 7 in FIG. 5A) is completed, the matching processing device 10 sets the matching code column to the initial value “0”. Is set (step S227). A match code (referred to as a completely unmatched code) indicating that no matching character string information has been found at any level is set for the record remaining as is. The complete mismatch code can be, for example, “999”, but may be any other character string or symbol, or the initial value “0” may be left as it is (in this case, , The complete mismatch code is “0”). As a result, for example, the databases A and B are in the state shown in FIG. 8, and the matching process is completed.

  Note that the matching code for each record of the databases A and B may be set in the original databases A and B read first, or in a new database obtained by duplicating the original databases A and B. May be set. Further, when a new database is created by duplicating the original databases A and B, one integrated database is created by integrating the contents of these databases A and B, and a match code is set for each record in the database. You may.

  As shown in FIG. 8, the matching code set in each record of the databases A and B is based on the character string information included in each of the databases A and B, in which segment the common character string information exists. (The position of the value “1” of the match code), and it is possible to easily read the match / mismatch status from the match code.

  Further, by expanding the match code, the number of records having the common character string information exists in the database A, and the number of records having the common character string information exists in the database B. It may be possible to grasp whether or not there is. For example, in the above example, two record Nos. 5, 6 and three record Nos. Nos. 4 to 6 have common character string information in the segments S1 to S3 as represented by the matching code “111”. To this match code “111”, a number is added so that the number in each of the databases A and B is known (for example, the number of records in the database A and the number of records in the database are arranged in this order). 23 ". Further, the user can easily grasp the number of records having the common character string information in the databases A and B by the expanded matching code.

  When a plurality of records (two or more records) having common character string information exist in the database, the number of records does not indicate the number of records as described above, but indicates that a plurality of records exist. Or just For example, a record in which only one record having common character string information exists in the database is represented by a value “1”, and a record in which two or more records exist has three or more records. May be represented by the value “2” even when the record of “1” exists. Thereby, "11" when the relationship between the numbers of records existing in the two databases is one-to-one, "12" or "21" when the relationship is one-to-many, and "22" when the relationship is many-to-many. It can be expressed as follows. It is sufficiently useful only to be able to grasp whether the relationship between the numbers of records existing in the two databases is one-to-one, one-to-many, or many-to-many.

  For example, in the above example, the match code “111: 23” is “111: 22” (the last “22” indicates that two or more records exist in the database A and two or more records exist in the database B). Is shown). FIG. 9 shows an example in which a match code represented by a value “2” is set for a record in which two or more records having common character string information exist in the database as described above. In FIG. 9, the complete mismatch code is set to "999: 99" to match the number of digits.

  Further, the matching processing apparatus 10 extracts only records in which a specific matching code is set from databases A and B in which matching data is set in each record, and displays only those records on a display screen. Or a new database containing only that record may be created and saved. For example, as shown in FIG. 9, in the example of the databases A and B in which the matching code to which the identification information of the single or plural records is added is set, the matching processing device 10 is configured to input a specific code from a user input or another device or program. When the output instruction of the match code “11_” is received, as shown in FIG. 7 and the record No. in the database B. The records 7 and 8 may be extracted and displayed with an identifier of the database in which each record exists. Records having the same matching code set in different databases A and B are likely to be related to each other, and by collectively displaying them on one display screen in this way, the user can determine whether the records match or mismatch. The situation can be easily grasped, and the work efficiency at the time of correcting a mismatched record can be improved.

  It should be noted that the flowcharts shown in FIGS. 2A to 2C can be extended to cope with the case where the matching process is performed in three or more databases. For example, when performing the matching process on three databases including the databases C in addition to the databases A and B, the maximum value vCMAX of the records of the database C is extracted in step S201, and the variable C1 of the number of records in the database C is extracted. Is incremented from 1 to vCMAX, in step S207, a loop process for searching for records common to databases A, B, and C is executed, so that combinations of records in each database can be completely compared. It is desirable to do. It should be noted that even when the matching process is performed in three or more databases, the above-described matching code can be used as it is. When information indicating the number of records is added to the matching code, the number of digits may be increased according to the number of databases. For example, the matching code “111: 121” (the last three digits are assigned to each of the three databases) (Corresponding to the number of existing records).

<Example>
The matching processing apparatus 10 according to the present invention can discover the difference between the two databases. Such differences include, for example, typographical errors or reading errors due to OCR (Optical Character Recognition) included in one of the two databases that should originally have the same information, database managers and data update timings. This is caused by the difference.

  Further, the character string information included in each record of the database to be processed by the matching processing apparatus 10 according to the present invention may be composed of, for example, only numbers, or may include kanji, kana, alphabets, symbols, and the like. Or the like. Further, in the present invention, it is assumed that a segment structure is set for character string information. In setting the segment structure, a hierarchical structure originally set hierarchically, such as address information, is used. Alternatively, the user of the matching processing apparatus 10 may segment the character string information by a unique method.

  Further, the type of database to be processed by the matching processing apparatus 10 according to the present invention is not particularly limited. For example, a database containing address information, a database containing network address information and domain information, a database containing personal information such as my number, a database containing customer management information, a database containing financial accounting information, a database containing product inventory information, etc. And various types of databases can be processed.

  In the following, a ledger database including address information and a map database including address information in association with image information representing a map or the like are prepared, and the matching process is actually performed by the matching processing device 10 according to the embodiment of the present invention. Examples will be described.

  In the ledger database and map database prepared in this embodiment, address information of the same area is stored in the same storage format. However, the managers and update times of the respective databases are different, and the records in the respective databases do not completely match. Further, as a result of addition or deletion of records in each database, the number of records may be different.

  In each record, address information having a hierarchical structure is stored as character string information. The character string information is an array of 23-digit numbers. The first three digits are municipalities, the next eight digits are large, the next four digits are small, the next five digits are lot numbers, and the last three digits are branch numbers. Respectively. For example, in the character string information “32200000004005800084005”, the first “322” is a municipal, the next “00000004” is a large letter, the next “0058” is a small letter, the next “00084” is a lot number, and the last “005” is a branch number. Is represented. In the reconciliation process in the present embodiment, based on the hierarchical structure of the address information, five segments of municipalities, large and small characters, lot numbers, and branch numbers are set, and a set of filters schematically shown in FIG. 5C is set. Is used.

  FIGS. 11 to 13 show the results of actually performing the abutting process under the above conditions by the abutting device 10 according to the embodiment of the present invention. FIGS. 11 to 13 show a state in which only records in which a match code specified by the user is set are extracted from each of the ledger database and the map database, and displayed on the display screen in a table format. These display states correspond to the display states of FIG. 10 described above.

  The value “1” of the field “type” in FIGS. 11 to 13 represents a record in the ledger database, and the value “2” represents a record in the map database. The fields "city", "large", "small", "lot number", and "branch number" in FIGS. 11 to 13 indicate each segment (each layer of address information) set in the character string information of each record. The field "CODE" contains the matching code. The address information, which is character string information, has five segments set therein, and the match code is also a five-digit number corresponding to each segment (the number of records in the ledger database and the number of records in the map database are also added). ).

  FIG. 11 illustrates a state in which records in which the match code “11 — 11” is set are extracted and displayed on the display screen. For example, the first and second records in the table of FIG. 11 are records in a ledger database and a map database, respectively, and these records have a correspondence. The first and second records in the table of FIG. 11 include character string information common to “municipalities”, “larger characters”, “lot numbers”, and “branch numbers” as indicated by the matching code “11_11”. On the other hand, small letters are different. Further, as indicated by “11” added to the end of the matching code, the first and second records of the table in FIG. 11 correspond one-to-one in the ledger database and the map database.

  FIG. 12 illustrates a state where a record in which the match code “11_1_” is set is extracted and displayed on the display screen. For example, the first and second records in the table in FIG. 12 are records in the map database and the ledger database, respectively, and these records have a correspondence. The first and second records in the table of FIG. 12 include character string information common to “municipalities”, “large characters”, and “lot numbers” as indicated by the matching code “11_1_”, while “small characters” , “Branch numbers” are different. Further, as indicated by “11” added to the end of the matching code, the first and second records of the table in FIG. 11 correspond one-to-one in the ledger database and the map database.

  FIG. 13 illustrates a state where a record in which the match code “111_1” is set is extracted and displayed on the display screen. For example, the first and third records in the table in FIG. 13 are records in the map database and the ledger database, respectively, and these records have a correspondence. The first and third records in the table of FIG. 13 include character string information common to “municipalities”, “large characters”, “small characters”, and “branch numbers” as indicated by the matching code “111_1”. On the other hand, the lot number is different. Further, as indicated by “11” added to the end of the matching code, the first and third records in the table in FIG. 13 correspond one-to-one in the ledger database and the map database.

  In addition, by using the fact that the address information in the map database is associated with the image information for displaying the map, referring to the map created and displayed based on the information stored in the map database. Then, the confirmation and correction of the abutting result according to the present invention may be performed.

  For example, the third and fourth records in the table of FIG. 11 are records in the map database and the ledger database, respectively, and these records have a correspondence. Specifically, the third and fourth records in the table of FIG. 11 are one-to-one records as indicated by the matching code “11_11: 11”, and include “municipalities”, “larger characters”, and “lot numbers”. , “Branch number” includes common character string information, but only “small letters” are different. More specifically, the third record (record in the map database) of the table in FIG. 11 has a value of “0170” in “lower case”, while the fourth record (record in the ledger database) is “ The only difference is that the value of "subscript" is "0172". From the above matching result, it can be inferred that there is an error in the value of “lower case” in one of the third and fourth records of the table in FIG. 11, but which of “0170” and “0172” It's not easy to identify if is wrong.

  In such a case, by using the information in the map database to display a map including the position to be checked for the matching result on the display screen of the display unit 40, the user who has referred to this map can use the ledger database or the map. It is possible to easily guess which record value of the database is wrong.

  FIG. 14 illustrates a state where a map around the address related to the third record (record in the map database) of the table in FIG. 11 is displayed on the display screen based on the information stored in the map database. Have been. Each record (including address information) in the map database represents a section including a specific address on the map, and in FIG. 14, a corresponding address (character string) is displayed together with an image showing each section. I have. For example, the third record in the table of FIG. 11 corresponds to the section A in the map of FIG. 14, and the character string “017200032-000” representing the address included in the third record is assigned to this section A. It is superimposed on the display.

  In the map of FIG. 14, both the section of “0170” and the section of “0172” with the value of “small font” are displayed, and the section of “0170” is colored and displayed so as to be visually emphasized. I have. In the example of FIG. 14, the section “0170” is hatched, but the hatching indicates coloring of a specific color. Similarly, the section “0172” may be colored and displayed (however, it is desirable that the section “0172” has a different color from the section “0172”).

  The user who refers to this map finds that the section A whose “small font” is set to “0170” is isolated within the area of “small font” “0172” due to the difference in coloring from the surrounding sections. It is possible to visually recognize that the state of the enclave has been reached. As a result, the user finds that the value “0170” (the value of the record in the map database) of the “small font” set in the section A is incorrect, and the value in the ledger database corresponding to the one-to-one obtained by the matching process. It can be inferred that the value “0172” of “subscript” included in the record of “No.” is likely to be a correct value, and the value of “subscript” of the third record in the table of FIG. 11 is changed from “0170” to “0172”. It can be determined that it should be changed.

  Here, after obtaining a record in which there is an inconsistency between the map database and the ledger database by referring to the result of the matching process, the map is displayed using the map displayed based on the information in the map database. Although the mode in which the inconsistency is confirmed has been described, as a different mode, after visually identifying a section (for example, an enclave, etc.) that visually feels a sense of incongruity with reference to a map displayed based on information in the map database, The record of the ledger database corresponding to the section may be checked with reference to the result of the matching process. Also, on a map displayed based on the information in the map database, information obtained by the matching process (for example, a matching code or character string information included in a record of the ledger database corresponding to one-to-one) is displayed. You may superimpose and display.

  In addition, the tables shown in FIGS. 11 to 13 include a feature that facilitates comparison of records by a user when displaying a result of the matching obtained by the matching process. Hereinafter, features for improving the visibility and convenience of the user in the table displaying the result of performing the matching process will be described.

  The tables in FIGS. 11 to 13 are characterized in that, for example, the sort display is performed based on the “municipalities”, “large characters”, “small characters” fields and the like. With this feature, records having common character string information are displayed on adjacent lines, and it is possible to improve the visibility of the user when comparing records having common character string information. Since records in different databases are displayed in adjacent rows, in the entire table, records in the ledger database (“type” = “1”) and records in the map database (“type” = “2”) ) Are displayed in a nested manner.

  The tables in FIGS. 11 to 13 are characterized in that the character string information in each record of the two compared databases (ledger database and map database) is displayed in different fields. More specifically, if the record is a record in the ledger database, the lot number and branch number of each record are displayed in the “building number” and “branch number” fields of the ledger database. In some cases, the information is displayed in the "lot number" and "branch number" fields of the map database. As described above, by separating the field for displaying the character string information included in the record of the ledger database from the field for displaying the character string information included in the record of the map database, the user can determine whether each record has the ledger database and the It will be possible to immediately determine which one of the map databases is included.

  In addition, as a result of the matching process according to the present invention, records that completely match in database A and database B and have a one-to-one correspondence (that is, in the above-described example, matching code “111: 11”). Or a record in which “11111: 11” is set), the matching result may not always be correct. Therefore, it may be possible to quantitatively verify whether or not the matching result is correct for the records determined to have a one-to-one correspondence.

  When confirming the result of matching between records determined to correspond one-to-one, for example, a common item included in each record of two compared databases (ledger database and map database) is used. It is possible to Hereinafter, description will be made with reference to FIGS. 15 and 16.

  FIG. 15 shows a state where six records (hereinafter, referred to as first to sixth records in order from the top) in which the matching code “11111: 11” is set and displayed on the display screen. Is shown. However, the “area” field included in each record is set in the table of FIG. The “area” field is set for each record in the ledger database and the map database, and the “area” field stores the area of the land (map area) related to the address information stored in each record. Have been. If the records correspond to one-to-one in the ledger database and the map database, the numerical value (land area) of the “area” field, which is a common item, should be the same. Utilizing this relationship, the records in the ledger database and the map database determined to have a one-to-one correspondence are further compared with each other in the "area" field. By judging that the one-to-one matching result is correct, it is possible to verify the matching result.

  When judging whether or not the information included in the common item is equal, for example, if the difference between the numerical values falls within a predetermined range (for example, several percent), the information included in the common item is substantially (That is, the one-to-one matching result is correct). This predetermined range can be freely set by the user.

  Among the first to sixth records shown in FIG. 15, the first and second records are character strings common to all of “municipalities”, “large characters”, “small characters”, “local numbers”, and “branch numbers”. The third and fourth records include character string information "20600021000000356002" common to all of "municipalities", "large characters", "small characters", "chiban", and "branch numbers". The sixth record includes character string information "20600021000000356003" common to all of "municipalities", "large characters", "small characters", "local numbers", and "branches". The first, third, and sixth records are records in the ledger database, and the second, fourth, and fifth records are records in the map database. In the table of FIG. 15, records having common character string information in all of “municipalities”, “large letters”, “small letters”, “local numbers”, and “branch numbers” are displayed in adjacent rows.

  As described above, the "area" field of each record is displayed in the table of FIG. For example, the first and second records are records determined to have a one-to-one correspondence in the matching process, but the values of the “area” field of the first record and the “area” field of the second record are both Since “946” is equal, it can be determined that the matching result that the first and second records correspond one-to-one is correct. The numerical value of the “area” field of the third record is “1470”, and the numerical value of the “area” field of the fourth record is “1452”. Since both numerical values are almost equal, the third and fourth records are It can be determined that the matching result of the one-to-one correspondence is correct. The numerical value of the “area” field of the fifth record is “1603”, and the numerical value of the “area” field of the sixth record is “810”. Since both numerical values are greatly different, the fifth and sixth records are different. It can be determined that the matching result of the one-to-one correspondence is an error.

  In the above description, for example, it is determined that the numerical value “1470” in the “area” field of the third record is substantially equal to the numerical value “1452” in the “area” field of the fourth record. Depends on freely settable criteria. For example, if the difference between the two numbers (here, “18”) is within 5% of one of the numbers or the average of both numbers, then if it is determined that both numbers are substantially equal, It is determined that the matching result that the third and fourth records have a one-to-one correspondence is correct. On the other hand, for example, if the difference between the two numerical values (here, “18”) is within 1% of the average value of either numerical value or both numerical values, it is determined that both numerical values are substantially equal. Is determined that the matching result that the third and fourth records correspond one-to-one is incorrect.

  Note that the above-described verification may be performed on records determined to correspond one-to-one, and the verification result may be set in each record. For example, it is possible to set a flag for a record for which the matching result that corresponds to one-to-one is determined to be incorrect or to change the color of the record when displaying it in a table and highlight it. is there.

  Further, as shown in FIG. 16, a scatter diagram in which numerical values in the “area” field are plotted may be created so that the user can visually determine whether or not the matching result is correct. FIG. 16 plots a number of points where the numerical values of the “area” fields of the two records determined to have a one-to-one correspondence are set as the X coordinate (horizontal axis) and the Y coordinate (vertical axis). . If the values of the "area" fields of the two records determined to correspond one-to-one are approximately equal, the plotted point is located on or near a straight line Y = X. You. Therefore, a point plotted far away from the Y = X straight line can be considered to represent an erroneous match result.

  In the scatter diagram shown in FIG. 16, for example, a straight line indicating the criterion set by the user may be displayed, and the color of a point exceeding the criterion and judged that the abutment result is incorrect is displayed. It may be changed and highlighted.

  In the above example, the case where the “area” field is used as an item common to the record in the ledger database and the record in the map database has been described, but the common field set in the records in a plurality of databases to be compared is described. As long as the item is, other items may be used. In the example of the ledger database and the map database, in addition to the area of the land, various items such as, for example, crops cultivated on the land, land marks indicating the use of the land, perimeter of the land, and the like are included. It is possible to use. In addition, here, the description is made with reference to the example of the ledger database and the map database. However, similarly, in other arbitrary types of data databases, common items set in the records in the databases to be compared are appropriately selected. It is possible to

  INDUSTRIAL APPLICABILITY The present invention is applicable to a technique for performing a reconciliation process on a plurality of databases each storing a plurality of records. Further, the present invention can process a database in which arbitrary types of data are stored, and is applicable to all database management technologies.

REFERENCE SIGNS LIST 10 Join processing device 11 Database reading unit 12 Join processing unit 13 Record comparison unit 14 Join result generation unit 15 Join result output unit 16 Segment setting unit 20 Storage medium 30 Operation input unit 40 Display unit

Claims (10)

A reconciliation processing device that performs reconciliation processing on a plurality of databases each storing a plurality of records including character string information,
A segment setting unit that sets a plurality of segments according to a predetermined rule for the character string information;
The character string information included in the record in the first database of the plurality of databases is compared with the character string information included in the record in the second database of the plurality of databases. A record comparing unit that detects a match or mismatch of character string information in a segment;
When a record including the common character string information in all of the compared segments is present in both the first and second databases, information indicating that the common character string information is included in all of the compared segments is stored. A matching result generation unit for associating with each of the records in the first and second databases including character string information common to all of the compared segments;
Butt processing device to have.   The matching processing apparatus according to claim 1, wherein the record comparison unit is configured to select one or a plurality of segments to be compared for matching or mismatching of the character string information from the plurality of segments. .   The record comparing unit first selects a predetermined number of the plurality of segments, detects the match or mismatch of the character string information in the predetermined number of the segments, and then matches or mismatches the string information. 3. The matching processing apparatus according to claim 2, wherein the matching processing apparatus is configured to detect a match or mismatch of the character string information while gradually decreasing the number of segments to be compared from the predetermined number.   The record comparison unit is configured to exclude, from the comparison targets, records in the first and second databases in which a match of the character string information in the corresponding segments has already been detected. The butting apparatus according to claim 2.   2. The matching result generation unit is configured to generate code information indicating which segment of the plurality of segments corresponds to all the compared segments including the common character string information. 5. The abutment processing apparatus according to any one of items 1 to 4. The code information includes information based on the number of records in the first database including common character string information in all of the compared segments, and the second database including common character string information in all of the compared segments. The matching processing apparatus according to claim 5, further comprising: information based on the number of records in the data.   Whether the information based on the number of records in the first and second databases is the number of records in the first and second databases themselves, or whether the number of records in the first and second databases is singular or plural The matching processing apparatus according to any one of claims 1 to 6, wherein the information is information indicating the following.   8. The method according to claim 1, wherein the character string information is address information, and the segment setting unit is configured to set the plurality of segments based on a hierarchical structure defined in advance in the address information. 9. A collision processing apparatus according to one of the above. A reconciliation processing method for reconciling a plurality of databases each storing a plurality of records including character string information,
A segment setting step of setting a plurality of segments according to a predetermined rule for the character string information;
The character string information included in the record in the first database of the plurality of databases is compared with the character string information included in the record in the second database of the plurality of databases. A record comparison step of detecting a match or mismatch of character string information in a segment;
When a record including the common character string information in all of the compared segments is present in both the first and second databases, information indicating that the common character string information is included in all of the compared segments is stored. Generating a matching result associated with each of the records in the first and second databases including character string information common to all of the compared segments.
Butting method. A match processing program for causing a computer to execute a match processing method of performing a match processing of a plurality of databases each storing a plurality of records including character string information,
A segment setting step of setting a plurality of segments according to a predetermined rule for the character string information;
The character string information included in the record in the first database of the plurality of databases is compared with the character string information included in the record in the second database of the plurality of databases. A record comparison step of detecting a match or mismatch of character string information in a segment;
When a record including the common character string information in all of the compared segments is present in both the first and second databases, information indicating that the common character string information is included in all of the compared segments is stored. Generating a matching result associated with each of the records in the first and second databases including character string information common to all of the compared segments.
A program for causing a computer to execute the matching processing method.