JP2021071789A - Application creation support device - Google Patents

Abstract

To simplify input to an electronic application without needing work of associating input items for which the same contents are to be input with each other as common input items.SOLUTION: An application creation support device 1 includes a learning model 12 in which the number of times of appearance of simultaneously appearing related item groups for each related item group, and an input support unit 10 which automatically transcribes an input item to an input item selected by an applicator 2. The input support unit 10 is established on prerequisite of input state of an electronic application 150, specifies a related item group having possibility of appearance on the prerequisite of the related item group or the like as a presumed related item group after extracting a related item group containing a selected input item as a related item group candidate, obtains a likelihood of the related item group candidate by using the number of times of appearance related to the assumed related item group, and uses the input value corresponding to a related item group candidate having the largest likelihood for the automatic transcription.SELECTED DRAWING: Figure 1

Description

The present invention is an invention for supporting the preparation of an electronic application form required for procedures related to life events.

Some life events, which are major events in daily life, do not require procedures such as application / notification to government agencies (for example, employment), but procedures such as application / notification to government agencies are required. There are many life events that are needed (eg marriage and moving). From now on, a life event means a life event that requires procedures such as application and notification to an administrative agency.

Citizens who had a life event had to go to the window of the administrative agency and receive a paper application form required for the procedure related to the life event that occurred from the staff at the window of the administrative agency. With the advent of society, a system has been developed that uses networks to carry out procedures related to life events.

In a system for performing procedures related to life events using a network, an electronic application form obtained by digitizing a paper medium application form is used instead of a paper medium application form as in the invention disclosed in Patent Document 1. Will be done. Since the electronic application form is different for each procedure, if there are multiple procedures related to the life event that occurred in the applicant, the applicant inputs the input values in all the input items provided for each of the multiple electronic application forms. It must be, and this system wants to simplify the entry into the electronic application form.

As a method to simplify the input to the electronic application form, if there is an input item (for example, the name of the applicant) to input the same input value in multiple different electronic application forms, input to this input item once. It has long been considered to be enough.

For example, the input support system disclosed in Patent Document 2 includes an item content input means that accepts input of item contents of a plurality of input items that require input by an operator, and a plurality of inputs that require input of item contents by an operator. For the common item associating means that associates the items for which the same contents should be input as common input items and the items that are associated as common input items by the common item associating means, the operator only needs to input the item contents once. It is provided with an input item control means for controlling the item content input means so as to be completed.

As a method of associating items for which the same contents should be input as common input items, a method of setting the same identifier for items that overlap in a plurality of applications such as a name and an address can be considered, but the invention disclosed in Patent Document 2 In, a table is prepared that associates items for which the same contents should be input as common input items.

Japanese Unexamined Patent Publication No. 11-328288 Japanese Unexamined Patent Publication No. 2005-190039

However, in order to simplify the input to the electronic application form, it is a very time-consuming task to associate the input items that should have the same contents as common input items, so the input items that should have the same contents should be input. It is desired that the input to the electronic application form can be simplified without the work of associating the items as common input items. Therefore, an object of the present invention is to simplify the input to the electronic application form without requiring the work of associating the input items to which the same contents should be input as the common input items.

The first invention that solves the above-mentioned problems is that the related item group that appears at the same time for each related item group obtained by grouping the input items having the same input value in the electronic application form prepared by the applicant. When a learning model that describes the number of occurrences and an input item are selected by the applicant, it is established when the input status of the electronic application form at this point is used as a precondition, and the input item selected by the applicant is included. After extracting the related item group as a related item group candidate from the learning model, it may appear when it is premised that the related item group candidate appears in addition to the input state of the electronic application form at this time. A certain related item group is specified as a hypothetical related item group, and the likelihood of the related item group candidate is obtained by using the number of occurrences related to the hypothetical related item group. The application form creation support device is provided with an input support unit that automatically posts an input value corresponding to the related item group candidate having the highest degree to an input item selected by the applicant.

The application form creation support device described in the first invention includes the input items selected by the applicant by using the learning model and the input state of the electronic application form as clues in order to solve the problem of the present invention. It is configured to calculate the likelihood of related item group candidates and automatically determine the input value to be automatically posted to the input item selected by the applicant.

Further, the second invention is the application preparation support device according to the first invention, in which the learning model forms a matrix in which the related item groups are arranged in rows and columns, and the components of the matrix correspond to the rows. Prerequisite that the input support unit refers to the learning model and the related item group candidate appears in the data having a structure that describes the number of occurrences of the related item group and the related item group corresponding to the column at the same time. The likelihood is calculated by dividing the sum of the number of occurrences of the hypothetical related item group by the sum of the number of occurrences of the hypothetical related item group so far in accordance with the above.

The second invention is an invention that embodies the structure of the learning data in the application form creation support device according to the first invention and the content of calculating the likelihood of the related item group candidate.

Further, the third invention is an input included in the electronic application form when one applicant completes all the preparations of the electronic application form necessary for the administrative procedure in the application form creation support device according to the first invention or the second invention. After analyzing the items and their input values and grouping all the input items included in the electronic application form created by the applicant into the related item groups, for each of the related item groups obtained by grouping, It is characterized by including a learning unit that increments the number of occurrences related to the related item group by one.

The third invention is an invention in which the learning function of the learning model is provided in the application form creation support device according to the first invention or the second invention.

According to the above-described invention, it is possible to simplify the input to the electronic application form without requiring the work of associating the input items to which the same contents should be input as the common input items.

The figure which showed the schematic structure of the system which arranged the application form creation support device. A block diagram of one server that constitutes the application form creation support device. The figure explaining an example of a plurality of electronic application forms stored in an electronic application form DB. The figure explaining an example of a learning model. The figure explaining the process of updating a learning model. The figure explaining an example of the learning model before the update. The figure explaining the process of automatically posting an input value to the input item selected by an applicant. The supplementary figure explaining the process of calculating the sum total of the number of occurrences of a hypothetical related item group so far. A supplementary diagram illustrating a process of calculating the total number of occurrences of a related item group when the precondition that a related item group candidate has appeared is followed.

From here, an embodiment according to the present invention will be described. The present embodiment is for facilitating the understanding of the present invention, and the present invention is not limited to the present embodiment. Unless otherwise specified, the drawings are schematic drawings drawn to facilitate understanding of the present invention.

FIG. 1 is a diagram showing a schematic configuration of a system in which the application form creation support device 1 according to the present invention is arranged. In FIG. 1, the application form creation support device 1 having a function of supporting the creation of an application form is a plurality of computer networks such as a local area network (LAN) and a wide area network (WAN: Wide Area Network). Is connected to the computer terminal 20 possessed by the applicant 2 who creates the electronic application form 150 via the network 3 including the above.

The electronic application form 150 created by using the application form creation support device 1 can be configured so that the computer terminal 20 can send the electronic application form 150 to the server of the administrative agency 4. In FIG. 1, the applicant 2 is connected to the computer terminal 20. The electronic application form 150 created by using the application form creation support device 1 is output to paper from the printer 21, and the paper application form 210 is submitted to the administrative agency 4. In the present embodiment, the computer terminal 20 is illustrated as a personal computer, but as the computer terminal 20, a smartphone or a tablet computer can be used.

The application form creation support device 1 is a device composed of one or a plurality of servers. FIG. 2 illustrates a block diagram of one server 5 constituting the application form creation support device 1. As illustrated in FIG. 2, one server 5 constituting the application form creation support device 1 has one or more processors 50 and various components connected to the processor 50 via the bus 53. FIG. 2 illustrates a memory 51 including a RAM 510 (Random Access Memory) and a storage system 511 and a network adapter 52 as components connected to the processor 50 via the bus 53, but the processor via the bus 53. The components connected to 50 are not limited to these.

The application form creation support device 1 illustrated in FIG. 1 has an electronic application form DB 15 (DB: DataBase), an application form search function 14, a learning model 12, and an application form creation support function in order to support the creation of the electronic application form 150. 13 is provided.

The electronic application form DB 15 is realized by using the storage system 511. The electronic application form DB 15 stores the electronic application form 150 for various procedures to be submitted to the administrative agency 4. The attributes given to the electronic application 150 registered in the electronic application DB 15 include the name of the procedure corresponding to the electronic application 150 and the name of the life event that requires this procedure. It is possible to search the electronic application form 150 using the search key.

The application form search function 14 is a function realized by a computer program that operates the processor 50. The electronic application form 150 including the life event name received from the computer terminal 20 as an attribute is searched from the electronic application form DB 15, and the search result is obtained. The list of the electronic application forms 150 included in the above is transmitted to the computer terminal 20. In the list transmitted by the application form search function 14 to the computer terminal 20, the electronic application form 150 included in the list can be individually selected. When the applicant 2 selects one of the electronic application forms 150 included in the list, the electronic application form 150 selected by the applicant 2 is transmitted from the application form search function 14 to the computer terminal 20, and the applicant 2 is the computer terminal. The electronic application form 150 can be input using 20.

The learning model 12 is a related item group that appears at the same time for each related item group obtained by grouping input items having the same input value in a plurality of electronic application forms 150 created by one applicant 2. It is the data which describes the number of appearances of. The structure of the learning model 12 may be arbitrary, but the learning model 12 of the present embodiment forms a matrix in which related item groups are arranged in rows and columns, and is divided into matrix components and related item groups and columns corresponding to the rows. It has a structure that describes the number of times the corresponding related item group appears at the same time.

Such a learning model 12 is prepared by finding an input item that is highly related to the input item selected by the applicant 2 and inputting the input value of the input item that is highly related to the input item selected by the applicant 2. This is so that the applicant 2 can automatically post to the selected input item. In the learning model 12, the number of occurrences in which the related item group corresponding to the row and the related item group corresponding to the column appear at the same time includes the input item selected by the applicant 2 using the number of occurrences. This is because the likelihood representing the likelihood that the related item group is established is obtained, and the likelihood can be used to determine the input value to be automatically posted to the input item selected by the applicant 2.

The application form creation support function 13 provides the applicant 2 with various functions for supporting the creation of the electronic application form 150. In the present embodiment, the function provided by the application form creation support function 13 to the applicant 2 is limited to the input support function for automatically posting the input value to the input item selected by the applicant 2, but other than this. The function may be provided to the application form creation support function 13. As a function other than the input support function provided by the application form creation support function 13 to the applicant 2, there is a function of guiding the entry procedure of the electronic application form 150 input by the applicant 2 and precautions in the application.

The application form creation support function 13 has a learning unit 11 and an input support unit 10 as functions related to the input support function provided to the applicant 2. Each of the learning unit 11 and the input support unit 10 is a function realized by a computer program that operates the processor 50.

The learning unit 11 of the application form creation support function 13 executes a process of updating the learning model 12 based on the creation result of the electronic application form 150. Specifically, when the learning unit 11 of the application form creation support function 13 completes the creation of the electronic application form 150 required for the administrative procedure by one applicant 2, the input items included in the electronic application form 150 and their inputs are entered. All the values are analyzed, and the input items included in the electronic application form 150 created by the applicant 2 are grouped into related item groups. Next, the learning unit 11 refers to the learning model 12 and increments the number of occurrences related to the related item group by one for each related item group obtained by grouping, so that the electronic application form 150 The learning model 12 is updated according to the creation result of. In the present embodiment, the process of incrementing the number of occurrences related to the related item group by one is the related item obtained by grouping each row corresponding to the related item group obtained by grouping. It is a process to increment the number of occurrences stored in the column corresponding to the group.

The input support unit 10 of the application form creation support function 13 is a function of executing a process of automatically posting an input value to an input item selected by the applicant 2. In addition to automatically inputting input values to the input items selected by applicant 2, the automatic posting also includes suggesting (displaying input candidates) the input values to the input items selected by applicant 2. Is done.

Specifically, the input support unit 10 of the application form creation support function 13 is established when the applicant 2 selects an input item and the input state of the electronic application form 150 at this time is used as a precondition, and the applicant The related item group including the input item selected by 2 is extracted from the learning model 12 as a related item group candidate. Next, the input support unit 10 of the application form creation support function 13 sets a related item group that may appear when it is premised that a related item group candidate appears in addition to the input state of the electronic application form 150. Specify as a hypothetical related item group, find the likelihood of the related item group candidate for each related item group candidate using the number of occurrences related to the hypothetical related item group, and input value corresponding to the related item group candidate with the highest likelihood. Is automatically posted to the input item selected by the applicant 2.

From here, the learning model 12 and the application form creation support function 13 will be described in detail. FIG. 3 is a diagram illustrating an example of a plurality of electronic application forms 150 stored in the electronic application form DB 15. FIG. 4 is a diagram illustrating an example of the learning model 12. FIG. 5 is a diagram illustrating a process of updating the learning model 12. FIG. 6 is a diagram illustrating an example of the learning model 12 before the update. FIG. 7 is a diagram illustrating a process of automatically posting an input value to an input item selected by the applicant 2. FIG. 8 is a supplementary diagram illustrating a process of obtaining the total number of occurrences of the hypothetical related item group so far. FIG. 9 is a supplementary diagram illustrating a process of obtaining the total number of occurrences of the hypothetical related item group when the precondition that the related item group candidate has appeared is followed.

The number of electronic application forms 150 stored in the electronic application form DB15 may be arbitrary, but here, in order to make the explanation easier to understand, the electronic application form 150 stored in the electronic application form DB15 is stored in the electronic application form 150a and the electronic application form. There are three types: 150b and electronic application form 150c. As illustrated in FIG. 3, as input items to be input by the applicant 2, the electronic application form 150a is provided with an input item 150a1 for inputting a name and an input item 150a2 for inputting a date of birth. Further, in the electronic application form 150b, in addition to the input item 150b1 for inputting the name and the input item 150b2 for inputting the date of birth, the input item 150b3 for inputting another name and the input item 150b4 for inputting another date of birth Is provided. Further, the electronic application form 150c is provided with an input item 150c1 for inputting a name and an input item 150c2 for inputting a date of birth.

In the learning model 12, an identifier (for example, "a1") unique to each input item is added to each input item provided in the electronic application form 150 stored in the electronic application form DB 15, and the identifier is used to electronically. Each of the input items provided in the application form 150 is managed.

In FIG. 4, the learning model 12 has a matrix shape. Rows and columns of the learning model 12 are provided for each related item group obtained by grouping the input items having the same input value in at least the electronic application form 150 created by one applicant 2. There is no duplication in the related item groups arranged in the rows and columns of the learning model 12, but the related item groups are not fixed. When a related item group that is not registered in the learning model 12 is obtained, a row and a column corresponding to this related item group are newly added to the learning model 12.

For example, in FIG. 3, the input values of the input items 150a1, the input items 150b1 and the input items 150c1 are the same, the input values of the input items 150a2, the input items 150b2 and the input items 150c2 are the same, and the input values are the same as the input items 150b3. It is also assumed that there is no input item whose input value is the same as that of the input item 150b4. In this case, the related item groups are the related item group (a1, b1, c1), the related item group (a2, b2, c2), the related item group (b3), and the related item group (b4). For example, when the learning model 12 does not have a row and a column corresponding to the related item group (b4), a new row and a column corresponding to the related item group (b4) are added to the learning model 12. The related item group (a1, b1, c1) means a related item group composed of three input items 150a1, input item 150b1 and input item 150c1.

In the matrix component of the learning model 12 in the form of a matrix, the number of occurrences in which the related item group corresponding to the row and the related item group corresponding to the column appear at the same time is stored. According to FIG. 4, for example, the related item groups appearing at the same time as the related item groups (a1, b1, c1) are the related item group (a1, b1, c1), the related item group (a2, b2, c2), and the related items. The item group (b3) and the related item group (b4). The number of occurrences of each is twice for the related item group (a2, b2, c2), once for the related item group (b3), and once for the related item group (b4). Note that the number of occurrences stored in the matrix component in which the related item group corresponding to the row and the related item group corresponding to the column are the same is the number of occurrences of this related item group.

The learning unit 11 of the application form creation support function 13 is a function of updating the learning model 12 based on the input result of the electronic application form 150. Referring to FIG. 5, the learning unit 11 of the application form creation support function 13 is included in at least the electronic application form 150 created by the applicant 2 when updating the learning model 12 based on the input result of the electronic application form 150. The input items and their input values are analyzed, and all the input items included in the electronic application form 150 created by the applicant 2 are grouped for each input item having the same input value to obtain a related item group (S1). ). If there is an input item that has not been entered, the input item that has not been entered does not belong to any related item group.

Next, the learning unit 11 of the application form creation support function 13 confirms whether there is a related item group that is not in the learning model 12 at this time in the related item groups obtained by grouping (S2). Only when there is a related item group that is not in the learning model 12 at this time, the learning unit 11 of the application form creation support function 13 adds rows and columns corresponding to this related item group to the learning model 12 (S3).

Next, the learning unit 11 of the application form creation support function 13 executes a loop process (L1) that is repeated for each related item group obtained from the electronic application form 150 created by the applicant 2. In this loop processing (L1), the learning unit 11 of the application form creation support function 13 obtains from the electronic application form 150 created by the applicant 2 in the line corresponding to the related item group to be the target of the loop processing (L1). A process of adding one number of occurrences stored in the column corresponding to the related item group is performed (L2), and when this loop process (L1) is completed, the process of updating the learning model 12 is completed.

The process of updating the learning model 12 will be described with reference to a specific example. The difference between the learning model 12 shown in FIG. 4 and the learning model 12 before update shown in FIG. 6 is that the value in the dotted line of the learning model 12 before update shown in FIG. 6 subtracts one value from FIG. It is in the point that it is the value that was set. For example, in the learning model 12 before update shown in FIG. 6, the first and second columns of the row corresponding to the related item group (a1, b1, c1) are "1", and the third and fourth columns are "1", respectively. It is "0".

In the electronic application form 150a illustrated in FIG. 3, it is assumed that the input value of the input item 150a1 is "Nippon Taro" and the input value of the input item 150a2 is "May 5, 1990". In the electronic application form 150b illustrated in FIG. 3, the input value of the input item 150b1 is "Nippon Taro", the input value of the input item 150b2 is "May 5, 1990", and the input value of the input item 150b3 is "Nippon Hanako". , It is assumed that the input value of the input item 150b4 is "March 3, 1991". Further, in the electronic application form 150c illustrated in FIG. 3, it is assumed that the input value of the input item 150c1 is "Nippon Taro" and the input value of the input item 150c2 is "May 5, 1990". In this case, the input items 150a1, the input items 150b1 and the input items 150c1 have the same input values as "Nippon Taro", and the input items 150b2, the input items 150b2 and the input items 150c2 have the same input values as "May 5, 1990". , Input item 150b3 and input item 150b4 have no input items having the same input value.

Therefore, in S1 of FIG. 5, the related item groups obtained from this input result are the related item group (a1, b1, c1), the related item group (a2, b2, c2), the related item group (b3), and the related item. It becomes a group (b4), and the loop processing of FIG. 5 is performed in each of the related item group (a1, b1, c1), the related item group (a2, b2, c2), the related item group (b3), and the related item group (b4). (L1) is executed.

For example, when the related item group (a1, b1, c1) is the target of the loop processing (L1), the related item group that appears at the same time as the related item group (a1, b1, c1) is the related item group (a1, b1). , C1), related item group (a2, b2, c2), related item group (b3) and related item group (b4). , The number of appearances stored in columns 3 and 4 will be added by one.

Similar processing is performed on the row corresponding to the related item group (a2, b2, c2), the row corresponding to the related item group (b3), and the row corresponding to the related item group (b4), which is illustrated in FIG. The learning model 12 is updated to the learning model 12 illustrated in FIG.

By referring to such a learning model 12, other related item groups that are strongly related to the related item group can be inferred. In the learning model 12 illustrated in FIG. 4, in the row corresponding to the related item group (a1, b1, c1), the other related item group having the highest number of occurrences is the related item group (a2, b2, c2). , When the related item group (a1, b1, c1) appears, it can be estimated that the related item group (a2, b2, c2) is also likely to appear.

The input support unit 10 of the application form creation support function 13 uses this to determine the input value to be automatically posted to the input item selected by the applicant 2. Referring to FIG. 7, when the input support unit 10 of the application form creation support function 13 determines the input value to be automatically posted to the input item selected by the applicant 2, the input state of the electronic application form 150 at this time is first determined. Is established as a precondition, and a related item group including an input item selected by the applicant 2 is extracted from the learning model 12 as a related item group candidate (S10).

Next, the input support unit 10 of the application form creation support function 13 executes a loop process (L10) to be executed for each related item group candidate extracted from the learning model 12. In the loop processing (L10), in order to obtain the likelihood of the related item group candidate, the input support unit 10 of the application form creation support function 13 first refers to the learning model 12, and the electronic application form 150 at this point. The related item group that may appear when it is premised that the related item group candidate to be the target of the loop processing (L10) appears in addition to the input state of (L11) is specified as the assumed related item group (L11). Next, the input support unit 10 of the application form creation support function 13 obtains the total number of appearances Q, which is the sum of the number of appearances of the specified hypothetical related item group so far (L12). Next, the input support unit 10 of the application form creation support function 13 obtains the total number of appearances P, which is the sum of the number of appearances of the assumed related item group when the precondition that the related item group candidate appears is followed (L13). Next, the input support unit 10 of the application form creation support function 13 divides the total number of appearances P by the total number of appearances Q to calculate the likelihood of the related item group candidate (L14).

When the loop processing (L10) is completed, the input support unit 10 of the application form creation support function 13 determines the related item group candidate having the highest likelihood as the related item group candidate to be used for automatic posting. (S11). Next, the input support unit 10 of the application form creation support function 13 determines the input value corresponding to the related item group candidate used for automatic posting as the input value used for automatic posting (S12), and then the input value used for automatic posting. Is automatically posted to the input item selected by the applicant 2 (S13), and the process of automatically posting the input value to the input item selected by the applicant 2 ends.

The process of automatically posting the input value to the input item selected by the applicant 2 will be described with reference to a specific example. According to S10 of FIG. 7, the input state of the electronic application form 150 at this time is required for the process of automatically posting the input value to the input item selected by the applicant 2. Here, the input state of the electronic application form 150 at this time is as follows.
(1) The input values of the input items 150a1 of the electronic application form 150a and the input items 150b1 of the electronic application form 150b are the same. (2) The input values of the input items 150a2 of the electronic application form 150a and the input items 150b2 of the electronic application form 150b are the same. (3) The input value of the input item 150b3 of the electronic application form 150b and the input value 150c1 of the electronic application form 150c are the same. (4) The input value is the same as the input item 150b4 of the electronic application form 150b. As a precondition, consider the case where the input item selected by the applicant 2 is the input item 150c2 of the electronic application form 150c. With reference to the learning model 12 illustrated in FIG. 4, the related item group candidates, which are related item groups including the input items selected by the applicant 2, which are established when the input state so far is used as a precondition, are related. There are three item groups (a2, b2, c2), related item groups (b4, c2), and related item groups (c2).

Related items Explain that the group candidates are the above three. With reference to the learning model 12 illustrated in FIG. 4, the related item groups including the input items (here, the input items 150c2) selected by the applicant 2 are the related item groups (a2, b2, c2) and the related item groups (a2, b2, c2). There are five groups: b4, c2), related item group (b2, c2), related item group (a2, c2), and related item group (c2). Consider the related item group that is established when the input status so far is used as a precondition. From the above input state (2), since the input item 150a2 and the input item 150b2 are included in the same related item group, the related item group in which the input item 150a2 and the input item 150b2 exist independently, that is, the related item group. The item group (b2, c2) and the related item group (a2, c2) are excluded from the related item group candidates, and the related item group candidates are the above three.

The loop processing (L10) of FIG. 7 will be described with the related item groups (a2, b2, c2) as related item group candidates. The process of L11 in FIG. 7 will be described. In L11 of FIG. 7, the related item group that may appear when it is premised that the related item group candidate appears in addition to the input state of the electronic application form 150 is registered in the learning model and is this. Become a related item group that does not violate the prerequisites.

According to the learning model 12 illustrated in FIG. 4 and the above input states (1) to (4), when the related item group (a2, b2, c2) is used as the related item group candidate, it is registered in the learning model. Since the related item groups (a1, b1, c1) are contrary to the input state (3), they are not assumed to be related item groups. Since the related item groups (a2, b2, c2) registered in the learning model are candidates for the related item group, they are assumed related item groups. The related item group (b3) registered in the learning model does not become a hypothetical related item group because it violates the input state (3). In this way, by determining whether or not the preconditions are violated for each related item group registered in the learning model, the hypothetical related item group finally becomes a related item group candidate (related item group). In addition to a2, b2, c2), there are related item groups (b4), related item groups (a1, b1), and related item groups (b3, c1).

The processing of L12 in FIG. 7 will be described. Assumptions that appear at the same time as related item group candidates The total number of occurrences Q, which is the sum of the number of occurrences of related item groups so far, is the value (number of occurrences) of the matrix component included in the row (or column) of the hypothetical related item group. Becomes the sum of.

In FIG. 8, the rows corresponding to the hypothetical related item group are not shaded, and the other rows are shaded. In this case, the total number of occurrences Q is described in "6", which is the sum of the number of occurrences described in the line corresponding to the related item group (a2, b2, c2), and in the line corresponding to the related item group (b4). Corresponds to "12" which is the total number of occurrences, "14" which is the total number of appearances described in the line corresponding to the related item group (a1, b1), and related item group (b3, c1). The value is "40", which is the sum of "8", which is the sum of the number of occurrences described in the line.

The process of L13 in FIG. 7 will be described. Assumptions when the precondition that related item group candidates have appeared The total number of occurrences P, which is the sum of the number of occurrences of related item groups, is the value of the matrix component (number of occurrences) corresponding to the assumed related item group in each row and column. It becomes the sum of. In FIG. 9, the matrix components corresponding to the hypothetical related item groups in each of the rows and columns are not shaded, and the other matrix components are shaded. According to FIG. 9, when the related item group (a2, b2, c2) is used as the related item group candidate, the total number of occurrences P is “17” and the likelihood is “0.43”.

The loop processing (L10) of FIG. 7 is also executed for related item group candidates other than the related item groups (a2, b2, c2). With the same contents, referring to the learning model 12 illustrated in FIG. 4, the related item group (b4, c2) and the related item group (c2) that are candidates for the related item group other than the related item group (a2, b2, c2). ), Let's calculate the likelihood. When the related item group (b4, c2) is used as the related item group candidate, the total number of appearances Q is "44", the total number of appearances P is "32", and the likelihood is "0.73". When the related item group (c2) is used as a related item group candidate, the total number of appearances Q is "54", the total number of appearances P is "35", and the likelihood is "0.65".

In S11 of FIG. 7, the related item group candidate having the highest likelihood is the related item group (b4, c2). In this case, in S12 of FIG. 7, the related item group candidate used for automatic posting is the related item group (b4). , C2). In S13 of FIG. 7, the input value corresponding to the related item group (b4, c2) is the input item selected by the applicant 2 in the input item included in the related item group candidate used for automatic posting (here, the input item). The input values have been input so far in the input items other than 150c2) (here, the input items 150b4), and the input values are automatically posted to the input items selected by the applicant 2.

1 Application form creation support device 10 Input support unit 11 Learning unit 12 Learning model 13 Application form creation support function 14 Application form search function 15 Electronic application form DB
150 Electronic Application Form 2 Applicant 20 Computer Terminal

Claims (3)

A learning model that describes the number of occurrences of the related item group that appeared at the same time for each related item group obtained by grouping the input items with the same input value in the electronic application form created by the applicant. ，
When the applicant selects an input item, the learning is established when the input state of the electronic application form at this time is used as a precondition, and the related item group including the input item selected by the applicant is used as a related item group candidate. After extracting from the model, the related item group that may appear when it is assumed that the related item group candidate appears in addition to the input state of the electronic application form at this point is specified as the assumed related item group. , The process of obtaining the likelihood of the related item group candidate using the number of occurrences related to the assumed related item group is performed for each related item group candidate, and the input corresponding to the related item group candidate having the highest likelihood is performed. An input support unit that automatically posts the value to the input item selected by the applicant, and
An application form creation support device characterized by being equipped with. The learning model forms a matrix in which the related item groups are arranged in rows and columns, and the number of occurrences of the related item group corresponding to the row and the related item group corresponding to the column appearing at the same time as a component of the matrix is calculated. With the data of the structure described
The input support unit refers to the learning model and sets the total number of appearances of the hypothetical related item group in the case of following the precondition that the related item group candidate has appeared. The likelihood is calculated by dividing by the sum of
The application form creation support device described in claim 1, which is characterized in that. When one applicant completes all the electronic application forms required for administrative procedures, the input items included in the electronic application form and their input values are analyzed, and all the input items included in the electronic application form created by the applicant are analyzed. After grouping the input items into the related item groups, each of the related item groups obtained by grouping is provided with a learning unit that increments the number of occurrences related to the related item group by one. The application form creation support device according to claim 1 or 2.

Images