JPH07225807A - Census register data generator - Google Patents

Abstract

PURPOSE:To efficiently perform the revision work and the itemizing work required for census register data generation by systemizing a series of processes to generate census register data base from family registers and using code information. CONSTITUTION:Image data inputted from a scanner 1 is stored in a family register data storage part 2 and is converted to code data by a character recognition part 3 and is temporarily stored in a family register master 4. Thereafter, a data processing part 5 edits code data based on data read from an description pattern definition file 6 to output the result to a census register data base 7, a display device 9, and a printer 10 with respect to editing for revision work and itemization. When census register data is inputted from a keyboard 8, code data stored in the census register master 4 is sent to the data processing part, and the editing work of revision and itemization is performed, and edited census register data may be outputted to the display device 9 and the printer 10.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system for creating family register data by converting a family register stored and managed on paper into information necessary for processing by an electronic computer.

[0002]

2. Description of the Related Art Conventionally, an OCR or a character recognition device has been used to convert image data into code data in input work. However, regarding the data setup work for document data such as a family register, it is necessary to decide whether to store the items described in the family register for the image data during the data setup work in the family register database instead of the original family register law. It was necessary to carry out rework work to judge in light of the above, and itemization work to divide the description in text format into items, and such data editing was not considered.

[0003]

The above-mentioned prior art does not consider the modification work and the itemization work during the data setup. The data input work, the modification work, and the itemization work are classified into three types of work. To perform combined data setup work requires the knowledge of a family register clerical worker, and it takes time for human work, it is difficult to find even if an error occurs, and data tends to be uneven. Occur. Further, there is a problem that it takes time because it cannot be performed by a simple mechanical work.

The object of the present invention is to improve such problems, to improve the efficiency of the remodeling work and the itemization work necessary for preparing the family register data, and to prepare the family register data including the input work, the remodeling work, and the itemization work. To provide a device.

[0005]

In order to solve the above-mentioned problems, the family register data creating device of the present invention inputs the family register data by means of the family register data input means for inputting the family register as image information, and stores the family register data in the family register data storage unit. Remember. Next, the family register data is character-recognized by the data conversion means and converted from image information to code information. Data for which characters cannot be recognized by the data conversion means (hereinafter referred to as reject data) is output using the output means, corrected by the correction data input means, and input. Next, the family register data, which is character-recognized by the data converting unit and converted into code information, and the family register data input by the correction data input unit are stored in the family register master. Remodeling work to select whether or not to register the family register data stored in the family register master in the family register database by data conversion means, and itemization work to itemize the described items for each element, and the family register database creation means This is achieved by creating a database.

In order to solve the above-mentioned problems, the family register data creating apparatus of the present invention sorts the family register data stored in the family register data storage unit into type character type family register data and non-type character type family register data by a sorting means. To do. Characters of the assorted type character type family register data are recognized by the data conversion means and converted from image information to code information. The sorted non-type character type family register data is output by the image data output means, and is input by the code data input means based on the image data. It is also achieved by storing in the family register master the family register data which is character-recognized by the data converting means and converted into code information, the family register data input by the correction data inputting means and the family register data input by the code data inputting means. .

[0007]

Among the above means, the family register input means functions to input the image information of the family register, and the data conversion means functions to recognize the image information of the family register data character by character and convert it into code information. To do. The output means functions to output the characters that have not been converted to code information by the data conversion means as image information on a form or display, and the correction data input means functions to input the image information output by the output means as code information. To do. Further, the data converting means functions to edit the data by performing the reworking work and the itemizing work on the code data of the family register stored in the family register master, and the family data creating means stores the code information in the family database. It has a function.

Further, the sorting means character-recognizes the family register data, and when the recognition rate is equal to or higher than a preset threshold value, it is type character type family register data, and when the recognition rate is lower than the threshold value, it is non-type character type family register data. Sort. The data conversion means functions to character-recognize the image information of the type character type family register data character by character and convert it into code information, and the output means outputs the type character type family register data which has not been converted into code information by character recognition. It has a function of inputting the type character type family register data output by the correction data input means output means as code data. The image data output means has a function of outputting non-type character type family register data as image data to a form and a display, and the code data input means has a function of inputting code data based on the image data output by the image data output means. .

As described above, by combining the three types of work, that is, the data input work from the entry of the family register data to the creation of the database, the rework work, and the itemization work,
Since a large amount of input work by a person can be made as light as possible and accurate and uniform data can be provided in a short time, the above-mentioned problems of the prior art can be solved.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a system configuration diagram according to the present invention. 1, 1 is a scanner, 2 is a family register data storage unit, 3 is a character recognition unit, 4 is a family register master, 5 is a data processing unit, 6 is a description pattern definition file, 7 is a family register database, 8 is a keyboard, 9 is a display, 10
Is a printer.

The outline of the operation of the family register data creating apparatus of this embodiment will be described below. The image data input from the scanner 1 is stored in the family register data storage unit 2, then sent to the character recognition unit 3, converted into code data, and temporarily stored in the family register master 4. After that, the data processing unit 5 edits this code data based on the data read from the description pattern definition file 6 for the editing work / itemization editing, and outputs it to the family register database 7, the display 9 and the printer 10. . In addition, by inputting family register data from the keyboard 8, the code data stored in the family register master 4 is sent to the data processing unit 5 for reworking and itemizing editing work, and then the edited family register data is displayed. It is also possible to output to 9 and the printer 10.

FIG. 3 is an example showing the format of a family register.
Generally, the format of the sentences written in the family register is fixed,
As shown in FIG. 3, a plurality of case sentences 302 that describe matters such as “marriage” and “birth” form a family register sentence 301. That is, the family register sentence is expressed as follows.

Family register sentence = “case sentence” “case sentence” ... “Case sentence” Using the family register sentence 301 as an example, the following is obtained.

Family register statement = "Showa pick-up date 4th day of birth in Taito-ku, Tokyo Registered on the 5th day of the same month by father registration""September 8th year of Showa 2nd year of Haruka Aozora and marriage registration 5th, Higashi-Ueno, Taito-ku, Tokyo Hachiban Hitachi Ichiro's family register is registered. In addition, the case sentence is identification information that is multiple pieces of information that can be extracted, such as the "birth date" and "notification date", and the incident type (birth, "notification registration", etc. , Marital status, divorce, death, transfer, etc.) and characteristic data used to specify a description pattern in which the order of description of identification items is patterned.

Incident sentence = “identity matter” “identity matter” “characteristic data” “identity matter” “identity matter” “characteristic data” ...
..Using the case sentence 302 as an example, it becomes as follows.

Incident sentence = "Showa-cho 4th month of the year", "Taito-ku, Tokyo", "Birth in", "Fiveth day of the same month", "Father", "Notification enrollment" In the above sentence, the characteristic data are "birth in" and " Report registration ”.

[0018] For example, the pattern data of the sentence "Birth date of birth in Taito-ku, Tokyo on the 4th day of the Showa era on the 5th day of the same year" is "(birth date) (place of birth) birth (report date) (report “Person) Notification”, and this case sentence is “(Identity matter) (Identity matter) (Characteristic data) (Identity matter) (Identity matter) (Characteristic data) (NL)”. Here, "NL" is an abbreviation for new line and indicates a line break.

FIG. 2 shows the flow of processing of the family register data creating apparatus of this embodiment. Family register input storage means 101
Inputs the family register 100 as image data through the scanner 1 and stores it in the family register data storage unit 2. The scanner 1 is, for example, an image scanner device, and the family register data storage unit 2 is, for example, an optical disc. With respect to the image data of the family register stored in the family register data storage unit 2, the data conversion means 1060 character-recognizes the image data of the family register and converts the image data into a corresponding character code character by character. The output unit 1063 prints on the printer 10 or displays it on the display 9 as a reject list 1064 that lists reject data that cannot be recognized.

FIG. 7 shows the print format of the reject list 1064. Reference numeral 700 is, for example, a heading in which "family registration number" is printed. Reference numeral 701 is information indicating the position of the reject character, for example, "what line".
702 is information of “what character”. For example, when the character of the reject data is the 9th character as shown in FIG. 7, character positions (8, 9, 10 ) Is displayed. Reference numeral 703 represents one sentence including characters of reject data displayed as image information. The result of recognizing the one sentence and converting it into code information is 704. In this case, “?” Is displayed as shown in FIG. 7 for a character that cannot be recognized at all, and for example, as a result of the character recognition, a plurality of recognized character candidates are selected and the above-mentioned character position 702 is displayed. It may be displayed.

Based on the characters of the image information 703, the reject data characters of the reject list 1064 are
The correct data is input by the correction data input unit 1065, and the correction data is overwritten on the code information of the reject data character of the code information 704. Moreover, the data conversion means 1 is also used for normal data in which reject data does not appear.
In order to confirm the result of character recognition by 060, it is possible to output the confirmation list 1062 to the form or the display 9 on the printer 10 by the output means 1061 and correct it by the correction data input means 1065 if there is an error. . Next, the code data normally recognized by the data conversion means 1060 and the code data input by the correction data input means are stored in the family register master 4. Next, it is logically checked whether the data stored in the family register master 4 has been correctly input (108).
This logical check 108 is performed, for example, for checking the record structure such as "whether there is a personal record of the main person or the first person", "whether there is one or more enrolled persons", or for the name such as "death, grudge" in the name. Checks to check for inconsistencies in the entered data, such as checking if there are characters that are rarely used, checking whether the name of the town where the domicile is domicile actually exists, checking whether the spouse is "husband" or "wife", etc. is there. When the logic check 108 is completed, the data conversion means 109 performs itemization work and rework work. Using the description pattern data stored in advance in the description pattern definition file 110, a process of editing the family register information stored in the family register master 4 into necessary words and phrases (details will be described later). After editing the phrase, a consistency check is performed to confirm the consistency of the edited data (111). This check includes checking the date items (for example, checking before and after the date) and checking the consistency between items (a spouse's husband and wife are paired, or if there is an adoption, there is an adoptive family record field, or a fetus recognition matter. Immediately before is a birth matter etc.). This consistency check 11
When step 1 is completed, the family register information edited by the data converting means 109 is edited by the family register database editing means 112 and stored in the family register database 7 composed of a plurality of databases.

FIG. 8 is a block diagram of one database. The data consists of an item name 801 and an item byte count 802. For example, the family register number 803 has 10 digits for the numeric item, and the father name 804 has 30 digits for the Japanese item.

The sorting means 103 of the character recognition unit 3 converts the image data of the family register into the type character type family register data 1
05 and non-type character type family register data 1040 may be sorted. In this case, in the sorting means 103, for example, when the threshold is 90%, the character recognition rate for one line of the family register data is 90% or more is the type character type family register data, and the character recognition rate is less than 90%. Items are sorted as non-type character type family register data. The non-type character type family register data is mainly handwritten family register, and the character recognition rate is much lower than that of type letters. As for the sorted non-type character type family register data 1040, the family register data 1042 is output as image data by the image data output means 1041 by the printer 10 or displayed on the screen of the display 9. The code data input means 1043 inputs code data based on the output image information. The sorted type character type family register data 105 is input to the data conversion means 1060 and processed as described above. The code data input by the code data input means 1043 is stored in the family register master 4.

Next, the data conversion means 109 will be described with reference to FIGS. 4 and 5. Description pattern definition file 6
The description pattern data registered in advance is extracted from the table and expanded in a table (400).

FIG. 6 shows the data structure of the description pattern definition file 6 and the internal table. The description pattern data is divided into a plurality of patterns for each case such as birth and marriage, and all the description pattern data for developing the case sentence is stored. For example, when the case type is birth 600, there are 10 types of description pattern data as shown in 601. Each of the described pattern data represents a combination of the characteristic data and the identification item, the plurality of characteristic data items are arranged in the order of appearance, and the plurality of identification items are arranged in the order of appearance. In the pattern 1 (602), "birth" and "notification enrollment" are the characteristic data 603, and the described pattern data 604 in this case is registered in the order of "birth date""place of birth""reportdate""reporter". Has been done.

Returning to FIG. 4 and continuing the description, next, the family register data is input from the family register master 4 (401). When a plurality of family register data has already been input, the family register data to be processed is taken out. When the family register data is input, matching processing is performed on the description pattern data developed in the table and the entered family register data. In this family register data, case sentence data on the family register is stored in the structure of one sentence and one record. In the matching process, first, in order to avoid matching with all registered description pattern data, the case type is specified by the characteristic data for specifying the case type of the described item in advance (40
3). At this time, if it cannot be uniquely determined (404 NO)
Performs matching with the characteristic data of the description pattern for each case type (405). At this time, the characteristic data of the description pattern of the incident type that is not the true incident type does not match the characteristic data of the target sentence among the plurality of identified incident types, so that the incident type can be uniquely determined. After the incident type of the described item is determined, the description pattern is specified next. The description pattern is specified by a matching process of comparing the input data with the characteristic data for specifying the description pattern in the order of development in the internal table for all the description patterns of the target incident type (406). When the matching process is completed (407),
Next, the case sentence is expanded using the description pattern data of the specified description pattern (408). An example of the development of the case sentence is shown below.

Input data: "March 1, 1993, born in Chiyoda-ku, Tokyo, father reported on 2nd day of the same month, sent from the same mayor on 3rd day of the same month." Description pattern: Birth on (date of birth) (place of birth) (date of notification) )
(Notifier) Report (Date of receipt) (Receiver) Sending Matches with the entry pattern data sent and develops the case sentence.

(A) The first information of the described pattern data is the "birth date", and the following information is the identification item.

Since "birth date" is "date", delimit by "day" and extract "March 1, 1993" (b) The following information is "place of birth" The next is the characteristic data.

..... Separated by feature data "Birth",
Extract "Chiyoda-ku, Tokyo" (C) The following information is the "report date", and the next is the identification item.

.............. Since the "report date" is "date", it is separated by "day" and "2nd of the same month" is extracted. (D) The following information is the "reporter" and the next is It is characteristic data.

...... Separated by feature data "report",
"Father" is extracted. (E) The next information is "the date when it was sent", and the next is the identification item.

.............. Since the "sent date" is "date", delimit by "day" and extract "3rd of the same month". (F) The following information is "receiver", Next is the characteristic data.

................................................................................................................ ||| ... The same information is separated by "Sending enrollment", (G) The following information is NL.

........ End After expanding to the NL (409), the case sentence data is set in the work area of the data processing unit (410).

The case sentence data set in the work area of the data processing unit and the sentence are used to restore the sentence, and the consistency with the input data is checked (411).
Matching or inconsistency is selected (412). If no inconsistency occurs, the case sentence data in the work area of the data processing unit is set in the save area of the data processing unit (413). In the case of inconsistency, for example, it is set as an exception description pattern (41
4).

Of the case sentence data set in the save area, the omitted case sentence data is restored according to the determined rule (415). (“Same month” is changed to “March”, etc.) The itemization of family register data is completed by the above method (416).

Next, the itemizing work of the family register data will be described with reference to FIG. The data processing unit 5 has a registration condition definition table in which a condition for determining whether the case type should be stored in the family register database is registered. Based on this condition definition table, whether the itemized family register data is the data to be registered for each case sentence is selected based on the case type (501). For example, case name "marriage""cognition"
Is subject to registration, but "death" and "divorce" are not subject to registration. It is determined based on the condition definition table whether or not the family register data to be registered is data to be registered based on the characteristic data (502). For example, the case sentence including the feature data of "cognition" is not registered. Next, it is determined whether or not the family register data to be registered is the case sentence that is continuing at present based on the case sentence after the case sentence to be registered and the condition definition table (503). For example, if the characteristic data of “marriage cancellation” is present in the data after the case sentence of the case type “marriage”, the case sentence of the “marriage” is not registered because it is not continued.

[0039]

According to the present invention, a series of processes including an inputting work for making a family register database from a family register, a remodeling work, and an itemizing work are systemized, and by using code information, it is manually performed. There is an effect that it is possible to create a family register database by editing family register data with a minimum of work.

When a large number of handwritten family registers and dirty family registers are mixed, the family register data is sorted into the type character type family register data which is easy to recognize characters and the non-type character type family register data which is difficult to recognize characters. A database can be created and a family register database can be created efficiently.

[Brief description of drawings]

FIG. 1 is a system configuration diagram according to the present invention.

FIG. 2 is a diagram showing a flow of processing of the family register data creation device of the embodiment.

FIG. 3 is a diagram showing an example of a format of a family register.

FIG. 4 is a diagram showing a flow of processing of itemization work.

FIG. 5 is a diagram showing a flow of processing of a rework work.

FIG. 6 is a diagram showing a structure of written pattern data.

FIG. 7 is a diagram showing an example of a reject list.

FIG. 8 is a diagram showing an example of a data structure of a family register database.

[Explanation of symbols]

1 ... Scanner, 2 ... Family register data storage unit, 3
... Character recognition unit, 4 ... Family register master, 5 ... Data processing unit, 6 ... Described pattern definition file, 7 ...
..Family register database

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Takuya Okamoto 1099, Ozenji, Aso-ku, Kawasaki-shi, Kanagawa Inside the Hitachi, Ltd. Systems Development Laboratory (72) Inventor Takehiko Kumata 1-6-27 Shinsuna, Koto-ku, Tokyo Hitachi, Ltd. Public Information Division (72) Inventor Shinichi Yokoi 1-6-27 Shinsuna, Koto-ku, Tokyo Stock Corporation Hitachi Ltd. Public Information Division (72) Inventor Yukio Sakano 1-chome Shinsago, Koto-ku, Tokyo 6-27 No. 27, Public Information Division, Hitachi, Ltd. (72) Inventor Akitoshi Sakamoto 1-16-5 Dogenzaka, Shibuya-ku, Tokyo Inside Hitachi Information Systems Ltd. (72) Inventor, Nobuaki Kamada Dogenzaka, Shibuya-ku, Tokyo 1-16-5 Hitachi Information Systems Co., Ltd. (72) Inventor Kensuke Sarai Dogen, Shibuya-ku, Tokyo 1-16-5 Hitachi Information Systems Co., Ltd. (72) Inventor Masao Nishizawa Dogenzaka Shibuya-ku, Tokyo 1-16-5 Hitachi Information Systems Co., Ltd. (72) Inventor Shinji Omura 1 Dogenzaka Shibuya-ku, Tokyo Chome 16-5 Hitachi Information Systems Co., Ltd. (72) Inventor Akiko Kikuchi 1-16-5 Dogenzaka Shibuya-ku, Tokyo Inside Hitachi Information Systems Co., Ltd. (72) Katsuhisa Shibata 1-16 Dogenzaka Shibuya-ku, Tokyo No. 5 In Hitachi Information Systems Co., Ltd. (72) Inventor Masaru Akatsuka 1-16-5 Dogenzaka, Shibuya-ku, Tokyo Within Hitachi Information Systems Co., Ltd.

Claims (3)

[Claims] 1. A family register book input means for inputting a family register as image information, a family register data storage section for storing the family register data input by the family register input means, and a family register data storage section for storing the family register data. Data conversion means for recognizing the family register data by character and converting image information into code information, a family register master for storing the data code-converted by the data transforming means, and data stored in the family register master. A family register data creating device comprising: a family register database creating means for creating a family register database. 2. The data conversion means, depending on whether or not the recognition rate of character recognition of the family register data stored in the family register data storage unit is a predetermined value or more, type character type family register data and non-type character, respectively. 2. The family register data creation device according to claim 1, further comprising means for sorting into type family register data. 3. A storage means is provided for storing a plurality of characteristic data such as "birth" and "notification registration", and a plurality of written pattern data representing a combination of identification items such as "birth date" and "notification date". The family register database creating means compares the family register data input from the family register book master with the data stored in the storage means to identify matching pattern data, and then identifies the entered family register data as characteristic data. Alternatively, the family register data creation device according to claim 1, wherein the items are divided by items of identification.