JP2013250631A - Information processor, information processing method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently change a program for determining propriety of application content for merchandise contract in accordance with new merchandise without requiring any merchandise knowledge.SOLUTION: An abstract syntax tree is generated by inputting a source code in which a check rule for determining propriety of application content with respect to merchandise on the basis of attributes constituting contract conditions of the merchandise is described for each existing merchandise. Then, a nest structure is analyzed by using the generated abstract syntax tree, and an order relation between the attributes of each existing merchandise is set, and priority of each attribute described in the source code is set on the basis of the order relation between the attributes. Then, on the basis of the priority set for each attribute and a matching state of the content of the attributes of each existing merchandise with the content of the attributes of new merchandise, similar merchandise which is the most similar to the new merchandise is determined from among the existing merchandise.

Description

  The present invention relates to an information processing apparatus, an information processing method, and a program.

When a customer wants to make a contract for a financial product or insurance product (hereinafter also simply referred to as “product”), for example, the customer's application details such as various contract details and personal information entered in the application form Is examined to see if it meets the contract conditions of the product that it is trying to contract. In addition to the large number of check items for such examinations, the types and number of products handled are large, so it takes a lot of time for human work.
Therefore, a new contract system for life insurance companies is known in which a basic examination is mechanically performed on the system (for example, see Non-Patent Document 1).

Masanori Honda, “Knowing Businesses and Systems in the Financial Industry, Part 10 Life Insurance (2) Knowing the Structure of Core Systems Essential to Efficient Contracts”, [online], nikkei BPnet, [April 19, 2012 Day search], Internet (URL: http://itpro.nikkeibp.co.jp/article/lecture/20070301/263653/?ST=lecture)

For example, when a new product is released as a financial product or an insurance product, the screening program is changed so that the suitability of application contents can be checked corresponding to the new product.
There are many similar products. For this reason, for example, if the source code is to be changed efficiently in consideration of maintainability, it is necessary to accurately identify similar products. For this reason, an engineer who changes the source code according to a new product needs to have sufficient knowledge about the product. However, in reality, it is difficult to reliably secure such engineers, which hinders the efficiency of source code modification work.

  The present invention has been made in view of such circumstances, and a program for determining the suitability of application contents for a product contract can be efficiently changed according to a new product without requiring product knowledge. The purpose is to.

  In order to solve the above-described problem, an information processing apparatus according to one aspect of the present invention describes a determination rule for determining whether or not an application content for a product is appropriate based on an attribute of the product for each existing product. The syntactic analysis unit that generates the abstract syntax tree by inputting the source code, and the nested structure is analyzed using the generated abstract syntax tree to set the order relationship between the attributes for each existing product A priority structure setting unit for setting a priority order for each attribute described in the source code based on an order relation between the set attributes, and a priority order set for each attribute And a similar product determination unit that determines a similar product most similar to the new product from the existing products based on the matching state between the content of the attribute of the existing product and the content of the attribute of the new product. That.

  Further, in the information processing apparatus of the present invention, based on the location where the determination rule for the similar product is described in the source code, the contents to be changed in the source code for the addition of the determination rule for the new product A change determination unit for determining is further provided.

  In the information processing apparatus according to the present invention, the nested structure analysis unit sets a weighting value for the attribute for each existing product based on the content of the conditional expression including the attribute in the source code. An order relationship is set, and the priority order setting unit sets a priority order for each attribute described in the source code based on a weighting value set for the attribute for each existing product, and the similar product determination unit Determines the existing product having the highest similarity determined based on the weighting value set in the attribute that matches the content of the new product and each of the existing products as the similar product.

  In the information processing apparatus of the present invention, the nested structure analysis unit sets the weighting value for the same attribute to be greatly changed when the same attribute is included in a plurality of conditional statements in the same hierarchy.

  In the information processing apparatus according to the present invention, the nest structure analysis unit sets the weighting value for the attribute to be small when there are a plurality of values indicating the authenticity of the attribute included in the conditional expression. .

  In the information processing apparatus of the present invention, the priority order setting unit sets the priority order based on a transition rule about the order relation between the attributes analyzed as the nested structure, and the similar product determination unit includes: As a result of comparing the contents of the attributes until the contents of the attributes do not match between the new product and each of the existing products in order from the highest priority, the number of attributes with the same content is the largest. A product is determined as the similar product.

  In the information processing apparatus according to the present invention, the nested structure analysis unit sets a weighting value for the attribute for each existing product based on the content of the conditional expression including the attribute in the source code. The order relation is set, and the priority order setting unit sets the priority order based on the transition rule. As a result, for the plurality of attributes having the same priority order, the weighting set for the plurality of attributes is set. Use values to set different priorities.

  Also, an information processing method as one aspect of the present invention includes inputting a source code in which a determination rule for determining the suitability of application contents for a product based on the attribute of the product is described for each existing product. A syntactic analysis step for generating an abstract syntax tree; and a nest structure analysis step for setting an order relationship between the attributes for each of the existing products by analyzing a nest structure using the generated abstract syntax tree; A priority order setting step for setting a priority order for each attribute described in the source code based on the order relation between the set attributes, a priority order set for each attribute, and the existing product A similar product determination step for determining a similar product most similar to the new product from the existing products based on the matching state between the attribute content and the new product attribute content. That.

  Further, the program as one aspect of the present invention inputs a source code in which a determination rule for determining the suitability of application contents for a product based on the attribute of the product is described for each existing product to a computer. A syntactic analysis step for generating an abstract syntax tree; and a nest structure analysis step for setting an order relationship between the attributes for each of the existing products by analyzing a nested structure using the generated abstract syntax tree; A priority order setting step for setting a priority order for each attribute described in the source code based on the order relation between the set attributes; a priority order set for each attribute; and the existing product Similar product determination that determines a similar product that is most similar to the new product from the existing products based on a match state between the content of the attribute of the new product and the content of the attribute of the new product It is intended for executing the step.

  As described above, according to the present invention, it is possible to efficiently change a program for determining the suitability of application contents for a product contract according to a new product without requiring product knowledge. .

It is a figure which shows the structural example of the source code processing apparatus in embodiment of this invention. It is a figure which shows the example of a relationship between the contract conditions and check rules of a new product and similar products. It is a figure which shows the example of contract conditions with a new product and several similar goods candidate. It is a figure which shows the example of the content of the check rule of the source code and goods in 1st this embodiment. It is a figure which shows the example of the content of the check rule of the source code and goods in 1st Embodiment. It is a figure which shows the example of the content of the check rule of the source code and goods in 1st Embodiment. It is a figure which shows the example of the abstract syntax tree in 1st Embodiment. It is a figure which shows the example of the information table and priority order information in 1st Embodiment. It is a figure for demonstrating the process of the similar goods determination which the similar goods determination part in 1st Embodiment performs. It is a figure which shows the example of a display mode about the change location and description change content in the source code which the change determination part determined in 1st Embodiment. It is a figure which shows the example of a process sequence which the source code processing apparatus in 1st Embodiment performs. It is a figure which shows the example of a process sequence for the similar goods determination which the similar goods determination part in 1st Embodiment performs. It is a figure which shows the operation example of the nest structure analysis part in 2nd Embodiment. It is a figure which shows the operation example of the priority order setting part in 2nd Embodiment. It is a figure which shows the operation example of the similar goods determination part in 2nd Embodiment. It is a figure which shows the example of a process sequence for the nest information generation which the nest structure analysis part in 2nd Embodiment performs. It is a figure which shows the process sequence example for the similar goods determination which the similar goods determination part in 2nd Embodiment performs.

<First Embodiment>
[Configuration of source code processing device]
FIG. 1 shows a configuration example of a source code processing device 100 that is an information processing device according to an embodiment of the present invention. Note that the source code processing apparatus 100 according to the present embodiment has a source code of a program that determines (examines) whether a customer's application content meets the contract conditions of the financial product when the customer contracts the financial product. Treat as a target.

  The source code processing device 100 of this embodiment includes a source code storage unit 101, a syntax analysis unit 102, a nested structure analysis unit 103, a priority order setting unit 104, a similar product determination unit 105, a new product data storage unit 106, and a change determination unit. 107, a display unit 108, a source code changing unit 109, and an operation unit 110.

  The source code storage unit 101 stores source code. This source code is a program for determining the suitability of a customer's application based on attributes that form the contract conditions of the product when the customer contracts for a financial product (hereinafter simply referred to as “product”) as described above. It corresponds to. In other words, this source code describes a check rule (determination rule) for each existing product for determining whether or not the application content for the product is appropriate based on an attribute forming the contract condition of the product. In the present embodiment, the programming language for describing the source code is not particularly limited.

  The syntax analysis unit 102 inputs the source code stored in the source code storage unit 101 and generates an abstract syntax tree.

  The nest structure analysis unit 103 sets an order relationship between attributes for each existing product by analyzing the nest structure using the generated abstract syntax tree generated by the syntax analysis unit 102.

  The priority order setting unit 104 sets a priority order for each attribute described in the source code based on the order relation between attributes set by the nested structure analysis unit 103. In addition, the priority order setting unit 104 in the first embodiment sets the priority order based on the transition rule about the order relation between attributes analyzed as a nested structure.

  The similar product determination unit 105 is most similar to the new product among the existing products based on the priority order set for each attribute and the matching state between the content of the attribute of the existing product and the content of the attribute of the new product. Determine similar products. Specifically, the similar product determination unit 105 according to the first embodiment compares the attribute contents until the attribute contents do not match between the new product and each of the existing products in order from the highest priority. The existing product having the largest number of attributes having the same content is determined as a similar product.

  The new product data storage unit 106 stores new product data indicating contract conditions for new products. The hardware corresponding to the source code storage unit 101 and the new product data storage unit 106 can be, for example, an HDD (Hard Disc Drive), a flash memory, or the like. Moreover, the hardware as the source code storage unit 101 and the new product data storage unit 106 can be made common.

The change determination unit 107 determines the contents to be changed in the source code in order to add the check rule for the new product based on the location where the check rule for the similar product is described in the source code.
The change determination unit 107 according to the present embodiment determines a portion to be changed in the source code (change portion) and a description change content (description change content) in the changed portion as the content to be changed in the source code. Here, the changed part is a line (step) in the source code or a position in the line. Further, the description change contents include addition or deletion of words or steps.

The display unit 108 displays the contents to be changed in the source code determined by the change determination unit 107 according to the control of the change determination unit 107.
Note that it is also conceivable that the change determination unit 107 determines, for example, only the changed part and displays the changed part on the display unit 108.

The engineer (user) confirms the content displayed on the display unit 108 and performs an operation on the operation unit 110 to change the source code so that the check rule corresponding to the new product is reflected.
The source code changing unit 109 changes the contents of the source code stored in the source code storage unit 101 in accordance with an operation for changing the source code performed on the operation unit 110.

  The operation unit 110 collectively indicates operation devices for an engineer to operate the source code processing apparatus 100, and includes, for example, a keyboard and a mouse.

[Basic concept of source code change]
Next, the basic concept of source code change according to the new product in this embodiment will be described. In the present embodiment, the case of dealing with examination work related to product application for financial products will be described as an example, but the target products are not limited to financial products.
FIG. 2A shows an example of a contract condition (specification) for a new product. The contract terms of this new product are “Product Type” “Government Bond”, “Product ID” “90”, “Release Year” “2012”, “Redemption Period” “10 years”. Yes.
It should be noted that items such as “product type”, “product ID”, “release year”, and “redemption period” in the above contract conditions are treated as product attributes in the present embodiment. In FIG. 2, the above four attributes are illustrated for convenience, but more attributes may be set for each product.

  In the present embodiment, when changing the source code according to the new product, a similar product similar to the new product is searched from existing products. Then, the contents of the portion where the check rule corresponding to the similar product is described in the source code are copied, and the copied content is changed corresponding to the contract content of the new product.

  FIG. 2B shows contract conditions for similar products extracted as being similar to the new product. The contract condition of this similar product is that the “product type” attribute is “government bond”, the “product ID” attribute is “75”, the “release year” attribute is “2012”, and the “redemption period” attribute is “ 5 years ".

  FIG. 2C shows an example of the contents of a check rule extracted from the source code corresponding to the similar product in FIG. According to FIG. 2 (c), the check rules for similar products are as follows: "Error if the trading institution is" window sales "", "Error if the area payment is" 3 times payment "", "Transaction institution Is “A bank” and the fee payment is “twice payment”, an error (nonconformity) condition is indicated. In FIG. 2C, the content of the check rule is described for the sake of understanding, but in the source code, this content is described in a predetermined programming language.

And corresponding to a new product, the content of the check rule of FIG.2 (c) is copied, and this content is changed according to the contract conditions of a new product. In FIG. 2 (d), as a result of changing the contents of FIG. 2 (c) corresponding to the new product, the second and third items “area payment is“ three times payment ”error”, “transaction The contents of “Error is in the case where the institution is“ A bank ”and the payment is“ twice payment ”” are deleted.
In this embodiment, for example, the change shown in FIG. 2D is reflected in the source code. Accordingly, it is possible to determine (examine) whether or not the new product contract application is appropriate based on the changed source code.

Here, when specifying a similar product corresponding to a new product, it is necessary not only to depend on the number of attributes whose contents match, but to match as many as many important attributes as possible.
This point will be described with reference to FIG. 3A shows an example of contract conditions for a new product, and FIGS. 3B, 3C, and 3D show similar product candidates α, β, An example of contract conditions for γ is shown.
In this case, the similar product candidate α shown in FIG. 3B has the “product type” attribute and the “redemption period” attribute content “government bond” and “10 years”, respectively, and FIG. Is consistent with new products.
As shown in FIG. 3C, the similar product candidate β has “product type” attribute and “purchasable person” attribute contents “government bond” and “individual”, respectively. ) Match new products.
In addition, as shown in FIG. 3D, the similar product candidate γ has “2012” and “10 years” as the contents of each attribute of “release year” and “redemption period”. ) Match new products.

Here, in the similar product candidates α, β, and γ shown in FIGS. 3B, 3C, and 3D, the contents of the two attributes coincide with the new product. However, the combination of the two attributes is different for similar product candidates α, β, and γ.
As described above, it is necessary that similar products have as many high-priority attributes as possible to match new products. Accordingly, the degree of similarity of these similar product candidates α, β, and γ with respect to the new product differs depending on the combination of attributes whose contents match the new product.
In this case, for example, the attribute having the highest importance is “product type”, and the attribute having the next highest importance is “redemption period”. Therefore, among the similar product candidates α, β, and γ, the most similar to the new product has a new combination of “product type” and “redemption period” attributes as shown in FIG. This is a similar product candidate α that matches the product. That is, in this case, the similar product candidate α is selected as the similar product.

For example, in order for engineers to identify similar products of new products in consideration of the importance of product attributes as described above, sufficient knowledge of products is necessary. Is difficult. Even if such human resources can be secured, since the number of existing products is enormous, it takes considerable time to finally identify similar products.
Therefore, the source code processing device 100 according to the present embodiment automatically determines a similar product corresponding to the new product as described below.

[Example of source code]
FIG. 4 shows an example of the source code 120 stored in the source code storage unit 101 of the source code processing apparatus 100. The source code 120 shown in this figure is an example in which check rules for three existing products are described. Note that the number of existing products whose check rules are described in the source code 120 is not particularly limited, and may be more.
The three products whose check rules are described in the source code 120 of FIG. 4 are “ID = 13”, “ID = 10”, and “ID = 20” for the attribute of “ID (product ID)”, respectively. Is. A block BK1 in FIG. 4 indicates a block in which a plurality of check rules corresponding to the product with “ID = 13” are described.
5 and 6 show the same source code 120 as in FIG.
A block BK2 in FIG. 5 indicates a block in which a plurality of check rules corresponding to the product with “ID = 10” are described.
A block BK3 in FIG. 6 indicates a block in which a plurality of check rules corresponding to the product with “ID = 20” are described.

[Processing example of source code processing device]
Hereinafter, an example of processing executed by the source code processing device 100 for determining similar products and determining contents to be changed in the source code will be described.
First, the syntax analysis unit 102 reads the source code 120 having the contents shown in FIGS. 4 to 6 from the source code storage unit 101. Then, by parsing the read source code 120, an abstract syntax tree is generated.

FIG. 7 shows an example of an abstract syntax tree 200 generated by parsing the source code 120 having the contents shown in FIGS.
In FIG. 7, the node nd1 is the root. In addition, the tree structure portion including the nodes nd2 to nd11 corresponds to the block BK1 shown in FIG.
Further, the tree structure portion including the nodes nd2, nd4, and nd12 to nd20 corresponds to the block BK2 in FIG.
Further, the tree structure portion including the nodes nd21 to nd27 corresponds to the block BK3 in FIG.

  Next, the nested structure analysis unit 103 receives the generated abstract syntax tree 200 as shown in FIG. 7 and analyzes the nested structure. Nesting refers to the nested structure of blocks in source code.

  The nested structure in the case of the abstract syntax tree 200 of FIG. 7 is divided into three tree structure portions corresponding to the blocks BK1, BK2, and BK3. In response to this, the nest structure analysis unit 103 generates nest information corresponding to the nest structure for each of the blocks BK1, BK2, and BK3, and outputs the nest information as a nest information table.

FIG. 8A shows an example of a nested information table 300 output corresponding to the abstract syntax tree 200 of FIG.
The nest information table 300 shown in this figure has a structure in which nest information is associated with a nest information ID. Nest information ID = 1 corresponds to the block BK1. Nest information ID = 2 corresponds to the block BK2. Nest information ID = 3 corresponds to block BK1.
In generating the nest information, the nest structure analysis unit 103 performs depth-first search, so that the attribute used for the determination in the IF statement (conditional statement) with the shallower nest is first and the IF with the deeper nest The order relation between attributes is set so that the attribute used for determination in the sentence is later.

  For example, by searching for the node by the depth-first search for the tree structure portion of the block BK1 in FIG. 7, the nest structure analysis unit 103 shows the attribute used for the determination in the IF statement having the shallowest nest at the node nd3. Recognize that it is a “product type”. Further, the nest structure analysis unit 103 recognizes that the attribute used for the determination in the next deeply nested IF statement is the “redemption period” indicated in the node nd6, and further determines in the next deepest nested IF statement. It is recognized that the attributes used for are “ID” and “purchaser” shown in the node nd8.

As a result, the nest structure analysis unit 103 generates nest information with the content “product type> redemption period> ID | purchaser” as shown as nest information with nest information ID = 1 in FIG. It should be noted that the symbol “|” in the expression indicating the order relationship indicates that a plurality of attributes exist in one conditional expression. That is, for the two terms described before and after the symbol “|”, it means that neither logical product nor logical sum is required.
Accordingly, the order relationship of the attributes in the block BK1 is “redemption period” next to “product type”, followed by “ID” and “purchaser” in parallel.

  Also, the nest structure analysis unit 103 searches the tree structure portion corresponding to the block BK2 with depth priority, so that the nest structure analysis unit 103 shows the nest information with nest information ID = 2 in FIG. As described above, nest information having the content “product type> redemption period> purchase amount” is generated. That is, the order relationship of the attributes in the block BK2 is “the redemption period” next to “product type”, and then “the purchase amount”.

  Further, the nest structure analysis unit 103 searches the tree structure portion corresponding to the block BK3 with depth priority, and as shown as the nest information of the nest information ID = 3 in FIG. Nest information having the contents of ID | sales office is generated. That is, the order relationship of the attributes in the block BK3 is that “ID” and “sales office” are arranged in parallel after “product type”.

  Next, the priority order setting unit 104 sets a priority order for each attribute that appears in the IF statement of the source code, based on the nested information table 300 shown in FIG. Here, in the nest information table 300, all attributes appearing in the IF sentence of the source code are extracted, and the order relation is set for each block.

Therefore, the priority order setting unit 104 applies a transition rule to the attribute order relationship shown for each nested information in the nested information table 300, and sets the priority order for every attribute shown in the nested information table 300. To do. Thereby, the priority order is set for every attribute appearing in the IF sentence of the source code.
Each of the nest information shown in FIG. 8A corresponds to each block, but one block corresponds to a check rule corresponding to one product. Therefore, each of the nest information indicates an order relationship of attributes in the corresponding product.

The priority order setting unit 104 outputs priority order information indicating the set priority order. FIG. 8B is an example of priority order information 400 indicating the priority order set based on the nest information table 300 of FIG.
In the priority order information 400 shown in FIG. 8B, the “product type” attribute has the highest priority, and then the “redemption period” attribute has the highest priority. Following the “redemption period”, the four attributes of “purchaser”, “ID”, “purchase amount”, and “sales office” are in the same priority order in parallel.

  In the priority order shown in the priority order information 400 of FIG. 8B, the attribute appearing in the IF sentence with a shallow nest is higher. In the source code 120, an IF statement including an attribute that is often used in common for a plurality of products is shallower. In other words, the attribute included in the IF statement is more important as the IF statement is nested more shallowly. Focusing on this point, the priority order setting unit 104 sets the priority order so that the priority order of the attributes included in the IF sentence with a shallow nesting becomes higher.

  Next, the similar product determination unit 105 displays the contents of the attributes of the existing products of “ID = 13”, “ID = 10”, and “ID = 20” in which the check rules are described in the source code 120 in FIG. Sorting is performed according to the priority order indicated by the priority order information 400 shown in b). The similar product determination unit 105 generates this sort result as an existing product attribute table.

  The existing product attribute table 500 in FIG. 9A includes the attribute contents for each product of “ID = 13”, “ID = 10”, and “ID = 20” in the priority order information shown in FIG. Shows the result of sorting according to the priority order indicated by.

  In the existing product attribute table 500, the attribute in the leftmost column has the highest priority, and becomes lower as it goes to the right. In addition, what is necessary is just to determine the arrangement order according to a predetermined rule, and to sort about the attribute which became the same priority and became parallel.

  The similar product determination unit 105 also generates a new product attribute table 600 shown in FIG. For this purpose, the similar product determination unit 105 reads new product data from the new product data storage unit 106. Then, the similar product determination unit 105 sorts the content of the new product attribute indicated by the read new product data according to the priority order indicated by the priority order information 400 in FIG. Generate.

  Next, the similar product determination unit 105 performs comparison with the content of the attribute indicated by the new product attribute table 600 in order from the attribute with the highest priority for each existing product. At this time, the similar product determination unit 105 ends the comparison process for each existing product when a mismatch result is obtained.

Specifically, the attribute having the highest priority in FIG. 9A and FIG. 9B is “product type”.
Therefore, the similar product determination unit 105 first compares the content of “product type” with FIG. 9B for the product of “ID = 10” in the first line of FIG. In this case, the contents of the “product type” match each other with “government bonds”.

  Therefore, the similar product determination unit 105 compares the contents of the attribute of the “redemption period” with the second priority in the first row of FIG. 9A and FIG. 9B. Also in this case, the content of the attribute of “redemption period” matches with “10 years”.

  Therefore, the similar product determination unit 105 further. The content of the attribute of “purchaser” in the next priority order arranged on the right side is compared between the first line of FIG. 9A and FIG. 9B. At this stage, the attribute of the “purchaser” is “corporate, individual” in the first line of FIG. 9A, whereas FIG. 9B is only “corporation”. , The contents are inconsistent. Therefore, the similar product determination unit 105 ends the comparison of the attribute contents for the product on the first line in FIG.

  Next, the similar product determination unit 105 compares the content of “product type” with FIG. 9B for the product with “ID = 13” in the second row of FIG. In this case, the contents of the “product type” attribute match each other with “government bonds”.

  Therefore, the similar product determination unit 105 compares the contents of the attribute of the “redemption period” with the second priority in the second row of FIG. 9A and FIG. 9B. In this case, the content of the attribute of “redemption period” is “5 years” in the second line of FIG. 9A, whereas “10 years” in FIG. 9B. The contents are inconsistent. Therefore, the similar product determination unit 105 ends the comparison of the attribute contents for the product on the second line in FIG.

  Next, the similar product determination unit 105 compares the content of “product type” with FIG. 9B for the product of “ID = 20” in the third row of FIG. In this case, the contents of the “product type” match each other with “government bonds”. In this case, the content of the “product type” attribute is “local bond” in the third line of FIG. 9A, whereas “government bond” in FIG. 9B. Is inconsistent. Therefore, the similar product determination unit 105 ends the comparison of the attribute contents for the product in the third row in FIG.

  FIG. 9C reflects the result of the comparison as described above in the existing product attribute table 500. Based on the comparison results so far, the similar product determination unit 105 indicates that the number of attributes whose contents match that of the new product is “2” for the product of “ID = 10” in the first row of FIG. Recognize Further, the product of “ID = 13” on the second line in FIG. 9A recognizes that the number of attributes whose contents match that of the new product is “1”. Further, the product of “ID = 20” on the third line in FIG. 9A recognizes that the number of attributes whose contents match the new product is “0”.

  As existing products with a higher number of recognized attributes share more attributes with higher priority (i.e., higher importance) with new products, You can see that the degree of similarity is high. Therefore, the similar product determination unit 105 obtains the number of attributes recognized for each existing product as described above, for example, as the degree of similarity, as shown in the rightmost column of FIG. 9C. That is, the similarity of the product of “ID = 10” on the first line in FIG. 9A is “2”, and the similarity of the product of “ID = 13” on the second line is “1”. The similarity of the product with “ID = 20” in the third row is “0”.

  Then, the similar product determination unit 105 determines a product having the maximum similarity obtained in this way as a similar product. In the case of FIG. 9C, the similar product determination unit 105 determines the product with “ID = 10” in the first row as a similar product.

The change determination unit 107 determines contents to be changed in the source code 120 in order to add a new product check rule based on the similar product determination result by the similar product determination unit 105.
For this purpose, the change determination unit 107 inputs the ID (product ID) of the product determined as the similar product by the similar product determination unit 105. Also, the source code 120 stored in the source code storage unit 101 is read. Then, the change determination unit 107 searches for a block in which the check rule for the similar product is described in the source code 120 using the ID of the product determined as the similar product.

  Specifically, the product determined to be a similar product in FIG. 9C is the product with “ID = 10”. In order to search the source code 120 for a block in which the check rule for the product with “ID = 10” is described, the change determination unit 107 searches for a block including “ID = 10” in the IF sentence, for example. As a result, the change determination unit 107 searches the block BK2 in FIG. 5 as a block in which the check rule for the product with “ID = 10” is described.

  Next, the change determination unit 107 reads the new product data from the new product data storage unit 106, and the change determination unit 107 checks the check rule described in the block BK2 and the new product check rule indicated in the new product data. And recognize the difference. In this case, the change determination unit 107 recognizes that only the product ID of the new product is not described as a difference in the check rule of the block BK2.

  Therefore, the change determination unit 107 in this case determines that the line “IF ID = 10”, which includes the ID attribute in the IF statement in the block BK2, is the changed portion according to the difference recognized as described above.

  Furthermore, the change determination unit 107 of the present embodiment also determines the description change content at the changed part. In this case, since “ID = 30” of the new product should be added to the IF statement of “IF ID = 10”, the change determination unit 107 sets the line “IF ID = 10” for the change part It is determined that the description should be changed so that “IF ID = 10 OR 30” and “OR 30” are added at the end.

Then, the change determination unit 107 causes the display unit 108 to display the above determination result (change location and description change content).
FIG. 10 shows the contents of the determination result displayed on the display unit 108. As shown in this figure, the source code 120 is displayed on the display unit 108, for example. In addition, in the IF statement of “IF ID = 10” which is the changed portion in the source code 120, “OR 30” is added and displayed after this to indicate that this “OR 30” is the description change content. Emphasize by highlight display ind.

The engineer who has seen the display shown in FIG. 10, for example, confirms the content displayed in this way, and then performs an operation for rewriting the source code 120 so that the check rule for the new product is reflected. To do. In response to this operation, the source code changing unit 109 changes the source code 120 stored in the source code storage unit 101.
Thus, in the present embodiment, for example, even if the engineer does not have sufficient product knowledge, the source code processing apparatus automatically determines a similar product corresponding to the new product. And based on the determination result of similar goods, the contents which should be changed in the source code 120 can be determined, and an engineer can be notified. Note that the display mode of the determination result of the change determination unit 107 is not particularly limited to that shown in FIG.

[Example of processing procedure]
The flowchart in FIG. 11 illustrates an example of a processing procedure executed by the source code processing device 100.
First, the syntax analysis unit 102 reads the source code 120 from the source code storage unit 101 (step S101). Next, the syntax analysis unit 102 performs syntax analysis of the read source code 120 to generate an abstract syntax tree 200 (step S102).

  The nest structure analysis unit 103 uses the abstract syntax tree 200 generated in step S102 to generate nest information indicating the order relationship of attributes for each block (that is, for each existing product) (step S103).

  The priority order setting unit 104 sets the priority order for all attributes described in the source code 120 from the nesting information for each block generated in step S103 (step S104).

  The similar product determination unit 105 sets the priority order for each attribute set in step S104, the content of the attribute for each existing product, and the content of the new product attribute indicated by the new product data read from the new product data storage unit 106. Based on the above, one similar product similar to the new product is determined from the existing products (step S105).

Based on the product ID of the similar product determined in step S105 and the new product data, the change determination unit 107 determines the changed location in the source code 120 and the description change content at the changed location (step S106).
Then, the change determination unit 107 causes the display unit 108 to display the change location and the description change content as the determination result together with the source code (step S107).

  Then, the source code changing unit 109 changes the source code 120 stored in the source code storage unit 101 in accordance with an operation for changing the source code 120 performed by the engineer on the operation unit 110 (Step S110). S108).

  The flowchart in FIG. 12 illustrates an example of a processing procedure for determining similar products executed by the similar product determining unit 105 as step S105 in FIG.

  First, the similar product determination unit 105 generates a new product attribute table 600 shown in FIG. 9B in advance (step S201). That is, the similar product determination unit 105 sorts the attributes of the new products in correspondence with the priority order set in step S104 of FIG.

  Next, the similar product determination unit 105 assigns “1” to the variable n indicating the row number in the existing product attribute table 500 shown in FIGS. 9A and 9C (step S202).

Next, the similar product determination unit 105 selects the n-th row in the existing product attribute table 500 (step S203), and further substitutes “1” for the variable m indicating the priority order (step S204).
In addition, the similar product determination unit 105 compares the contents of the m-th priority attribute in the n-th row of the existing product attribute table 500 and the new product attribute table 600 (step S205).

  The similar product determination unit 105 determines whether or not the contents of both attributes match as a comparison result in step S205 (step S206). If they match (step S206—YES), the similar product determination unit 105 increments the variable m (step S207), and returns to step S205. Thereby, the comparison about the content of the attribute of the next priority order is continued.

  On the other hand, when the contents do not match (step S206—NO), the similar product determination unit 105 determines the value of the variable m with respect to the similarity D for the product in the nth row of the existing product attribute table 500. Is substituted (step S208). That is, the similar product determination unit 105 obtains a value corresponding to the number of attributes determined to match as the similarity D in accordance with the description of FIG.

Next, the similar product determination unit 105 determines whether or not the variable n is the maximum value (step S209). In correspondence with the example of FIG. 9, the maximum value of the variable n is “3”.
When the variable n is not the maximum value (step S209—NO), there is still a row of the existing product attribute table 500 that has not been compared with the attribute contents. Therefore, in this case, the similar product determination unit 105 increments the variable n (step S210), and returns to step S203 to compare the attribute contents for the next line in the existing product attribute table 500.
On the other hand, when the variable n is the maximum value (step S209—YES), the attribute contents are compared for all the rows of the existing product attribute table 500. Therefore, the similar product determination unit 105 in this case determines an existing product corresponding to the row having the maximum similarity D as a similar product (step S211).

<Second Embodiment>
[Operation example of nest structure analysis unit, priority order setting unit, similar product determination unit]
Next, a second embodiment will be described. The overall configuration of the source code processing apparatus 100 in the second embodiment may be the same as that shown in FIG. Further, the overall processing procedure executed by the source code processing device 100 in the second embodiment may be the same as that shown in FIG.

The first embodiment and the second embodiment are different in the priority order setting method for attributes. In the first embodiment, the transition order is applied to the order relation of attributes for each block analyzed by the nested structure analysis unit 103 to set the priority order for each attribute.
On the other hand, in the second embodiment, the nest structure analysis unit 103 sets a weighting value for the attribute in each nest information, not depending on the depth of nest, and the attribute for each block according to this weighting value. Set the order relationship. Then, the priority order setting unit 104 sets the priority order for each attribute based on the attribute weight value for each nest information. Furthermore, the similar product determination unit 105 determines an existing product having the highest similarity obtained based on the weighting value set in the attribute that matches the content of the new product and each of the existing products as the similar product. .

[Example of setting weight values for attributes]
FIGS. 13A and 13B respectively show setting examples of weight values for attributes when the nest structure analysis unit 103 in the second embodiment generates nest information.
First, in FIG. 13A, three conditional statements (IF statements) of conditional statements L1, L2, and L3 are described in the same hierarchy in the source code 120. Among these, the conditional statement L1 includes an attribute of “ID (product ID)”, but the conditional statements L2 and L3 both include an attribute of “product type”. That is, the “product type” attribute is included in a plurality of conditional statements in the same hierarchy.

  Thus, when the same attribute is included in a plurality of conditional statements in the same hierarchy, it can be said that this attribute has high importance. Therefore, when the same attribute is included in a plurality of conditional statements in the same hierarchy, the nested structure analysis unit 103 sets so as to greatly change the weight value of this attribute.

  As an example, in the case of FIG. 13A, it is assumed that the default value of the weighting value is “0.4”. In the source code 120 of FIG. 13A, since only the conditional statement L1 includes the attribute “ID” in the same hierarchy, the nested structure analysis unit 103 sets “ID” as a default to the attribute “ID”. A weight value of “0.4” is set.

On the other hand, the two conditional statements L2 and L3 are both in the same hierarchy and include the same “product type” attribute. Therefore, for the attribute of “product type”, the nested structure analysis unit 103 sets, for example, “0.8” which is twice as large as “0.4” as a weighting value larger than the default “0.4”. .
In setting the weight value to be greatly changed in this way, the weight value may be changed to increase as the number of the same attributes existing in the same hierarchy increases.

In FIG. 13B, two conditional statements (IF statements) of conditional statements L4 and L5 are described in the same hierarchy in the source code 120. Among these, the true value for the “product type” attribute included in the conditional statement L4 is only one “government bond”.
In contrast, in the conditional statement L5, as described as “IF ID = 10 OR 20 OR 30”, the true values for the attribute of “ID” in the conditional expression are “10”, “20 3 of “30”. That is, there are a plurality of true values in the conditional statement L5. As described above, when there are a plurality of true values for the attribute included in the conditional expression, the nest structure analyzing unit 103 sets the weighting value for the attribute to be small.

As an example, the nested structure analysis unit 103 sets a default weight value of “0.4” for the “product type” attribute included in the conditional statement L4.
On the other hand, for the attribute of “ID” included in the conditional statement L5, for example, “0.2” that is 1/2 times “0.4” as a weighting value lower than the default “0.4”. Set. In setting the weighting value to be small as described above, the weighting value may be changed so as to decrease as the number of true values increases.

  Further, depending on the content of the conditional expression, the nested structure analysis unit 103 may set the weighting value for the attribute to be small when there are a plurality of false values for the attribute included in the conditional expression. . That is, the nest structure analysis unit 103 sets the weighting value for the attribute to be small when there are a plurality of values indicating the authenticity of the attribute included in the conditional expression of the source code.

  Note that when the nest structure analysis unit 103 in the second embodiment sets the order relationship between attributes for each existing product, the method of FIG. 13 (a) alone may be employed, or FIG. ) Only, or the methods shown in FIGS. 13A and 13B may be used in combination.

  Then, the priority order setting unit 104 in the second embodiment uses the weighting value, which is the basis for setting the order relation between attributes for each existing product, to set the priority order for each attribute described in the source code 120. Set. Specifically, the priority order may be set in the order of the attribute having the highest weighting value.

However, in the second embodiment, particularly when the method of FIG. 13B is employed, there is a possibility that the weight values set for the same attribute are different between the nest information in the nest information table 300.
As a specific example, in FIG. 14A, the weight value K for the attribute of “product type” in the nested information with the nested information ID = “1” and “2” is “1.0”. The weighting value K for the attribute of “product type” with nest information ID = “3” is “1.2”.

  Further, the weighting value K for the attribute of “redemption period” in the nest information of the nest information ID = “1” and “2” is “0.8”, whereas the “redemption” of the nest information ID = “3”. The weighting value K for the “period” attribute is “0.5”.

  The weighting value K for the attribute “ID” in the nest information with the nest information ID = “1” is “0.3”, whereas the weighting value K for the attribute “ID” with the nest information ID = “3”. The value K is “0.4”.

  In such a case, for example, one regular weighting value is determined for the attribute according to a predetermined rule. For example, the priority order setting unit 104 can determine, for example, one of a mode value, a maximum value, and a minimum value among a plurality of weight values as a normal weight value. Alternatively, the priority order setting unit 104 can calculate an average value of a plurality of weight values and determine the average value as a normal weight value.

FIG. 14B shows the priority when the priority order setting unit 104 sets the priority order according to the normal weight value obtained by calculating the average value corresponding to the nest information table 300 of FIG. The order information 400 is shown.
In this priority order information 400, the priority order is set in the order of “product type”, “redemption date”, “purchase amount”, “sales office”, “ID”, “purchaser”. The “product type” weight value K is “1.1”, the “redemption period” weight value K is “0.7”, the “purchase price” weight value K is “0.6”, and “sales office”. The weighting value K is “0.4”, the “ID” weighting value K is “0.35”, and the “purchaser” weighting value K is “0.3”.

And the similar goods determination part 105 in 2nd Embodiment calculates | requires the similarity with the new goods about the existing goods as follows.
FIG. 15A shows a new product attribute table 600 similar to FIG. 9B. FIG. 15B shows an example of an existing product attribute table 500 generated by the similar product determination unit 105 according to the priority order set by the priority order setting unit 104. Note that the weighting value K for each attribute shown in this figure is different from that shown in FIG. 14B for convenience of explanation. In the example of FIG. 15B, “product type” (weighting value K = 0.8), “redemption period” (weighting value K = 0.6), “buyer” (weighting value K = 0.4). ), “ID” (weighting value K = 0.3)...

In addition, the similar product determination unit 105 in the second embodiment specifies an attribute whose content matches that of the new product attribute table 600 for each row (existing product) in the existing product attribute table 500. And the similarity for every line (existing goods) in the existing goods attribute table 500 is calculated | required by adding the weighting value of the attribute with which the content corresponded.
In the case of FIG. 15, as indicated by hatching in FIG. 15B, the contents of two attributes of “product type” and “redemption period” are displayed in the first line of the existing product attribute table 500. , “JGB” and “10 years” respectively match the new product attribute table 600. In this case, since the weight value of the attribute “product type” is 0.8 and the weight value of the attribute “redemption period” is 0.6, the similarity is 1.4.
Further, in the second line of the existing product attribute table 500, the contents of the two attributes “product type” and “purchaser” are “government bond” and “individual”, respectively, which are the same as the new product attribute table 600. I'm doing it. Therefore, the similarity corresponding to the second line of the existing product attribute table 500 is 1.2.
In the third line of the existing product attribute table 500, only the content of the “redemption period” attribute is “10 years” and matches the new product attribute table 600. Therefore, the similarity corresponding to the third line of the existing product attribute table 500 is 0.6.

  Thus, after calculating | requiring a similarity, the similar goods determination part 105 determines a thing with the maximum similarity as a similar goods. That is, in the case of FIG. 15, it is determined that the product with “ID = 10” in the first line of the existing product attribute table 500 is a similar product.

[Example of processing procedure]
The flowchart in FIG. 16 illustrates an example of a processing procedure for generating nest information that is executed as the step S103 in FIG. 11 by the nest structure analysis unit 103 in the second embodiment.

  The nest structure analysis unit 103 first selects a root node in the abstract syntax tree 200 (step S301).

In addition, the nested structure analysis unit 103 determines whether or not the selected node is an IF statement (conditional statement) (step S302).
If it is an IF statement (step S302—YES), the nested structure analysis unit 103, according to the content of the conditional expression of the IF statement, as described in FIG. 13, weighting values for attributes included in the IF statement Is set (step S303). Then, the nested structure analysis unit 103 holds the attribute included in the conditional expression in association with the weighting value set in step S303 (step S304), and then proceeds to step S305.

  If it is not an IF statement (step S302-NO), and after executing step S304, the nested structure analysis unit 103 determines whether there is an unsearched child node among the child nodes of the currently selected node. Is determined (step S305).

  When there are no unsearched child nodes (step S305-NO), that is, when the node at the deepest level is reached, the nested structure analysis unit 103 associates all the currently held attributes with these attributes. The weighting value is output as one nest information (step S306). In this way, each time the search is performed up to the deepest hierarchy node, the nest information is sequentially output one by one, and the nest information table 300 is generated from the nest information.

  Next, the nest structure analysis unit 103 deletes the attribute output in step S306 and the weight value data associated therewith (step S307).

  Next, the nest structure analysis unit 103 changes the selection to the parent node of the node that has been selected so far (step S308), and has not yet searched among the child nodes that have the selected parent node as a child. It is determined whether or not something exists (step S309).

  When there is an unsearched child node (YES in step S309), the nest structure analysis unit 103 selects one unsearched child node (step S310), and the process returns to step S302. Accordingly, the process proceeds to processing for generating nested information corresponding to another block (that is, another existing product).

  On the other hand, when there is no unsearched child node (step S309-NO), the nest structure analysis unit 103 further determines whether the parent node selected in step S307 is the root of the abstract syntax tree 200 (step S311). ).

If it is not the root of the abstract syntax tree 200 (step S311-NO), the nested structure analysis unit 103 returns to step S308 to further select the parent node of the parent node selected in the previous step S308.
If it is the root of the abstract syntax tree 200 (step S311-YES), all nodes in the abstract syntax tree 200 are searched, and nesting information corresponding to all blocks (existing products) is obtained. . In this case, the nested structure analysis unit 103 ends the search for the abstract syntax tree 200 so far.

In the description so far, weighting is performed on the attribute included in the IF sentence. However, for example, an attribute included in another conditional sentence such as an IF-ELSE sentence may be targeted.
Also, the nest structure analysis unit 103 in the first embodiment can generate the nest information table 300 shown in FIG. 8A by performing a search process according to FIG. However, in the first embodiment, it is not necessary to set the weighting value in step S303, and in step S304, only the attributes included in the conditional expression need be retained. Further, in step S306, it is only necessary to output the nesting information based only on the retained attributes, and in step S307, only the already output attributes need be deleted.

  The flowchart in FIG. 17 illustrates an example of a processing procedure for similar product determination executed by the similar product determination unit 105 in the second embodiment as step S105 in FIG.

  First, the similar product determination unit 105 generates in advance a new product attribute table 600 shown in FIG. 9B, for example (step S401).

  Next, the similar product determination unit 105 assigns “1” to the variable n indicating the row number in the existing product attribute table 500 shown in FIGS. 9A and 9C (step S402).

Next, the similar product determination unit 105 selects the n-th row in the existing product attribute table 500 (step S403).
And the similar goods determination part 105 compares the content for every attribute with the existing goods and new goods of the nth line (step S404). Then, the similar product determination unit 105 calculates the similarity D with the new product for the existing product in the n-th row based on the comparison result in step S404 (step S405). As described with reference to FIG. 15, the similarity D can be obtained by adding the weight values of the attributes whose contents match.

Next, the similar product determination unit 105 determines whether or not the variable n is the maximum value (step S406).
When the variable n is not the maximum value (step S406—NO), the similar product determination unit 105 increments the variable n (step S407), and then returns to step S403, whereby the next line in the existing product attribute table 500 is displayed. Compare the attribute contents for.
On the other hand, when the variable n is the maximum value (step S406—YES), the similar product determination unit 105 determines an existing product corresponding to the row having the maximum similarity D as a similar product (step S408).

[Modification]
As a modification of the present embodiment, for priority order setting, the first embodiment and the second embodiment may be combined as follows.
That is, the nested structure analysis unit 103 according to the modification sets the order relationship of attributes by associating weight values according to the second embodiment.
In addition, the priority order setting unit 104 first sets the priority order for each attribute by the transition rule. As a result of setting the priority order based on the transition rule, for the plurality of attributes having the same priority order, a different priority order is set using the weighting values set for the plurality of attributes. Is. That is, for a plurality of attributes having the same priority order set based on the transition rule, different priority orders are determined again according to the weighting values set for the plurality of attributes.
At this time, as described with reference to FIG. 14, when a plurality of weighting values set for the same attribute exist between the nest information, 1 is obtained by a predetermined calculation such as obtaining an average value. One regular weight value may be determined.

  In the description of the embodiments so far, the financial product is taken as an example, but it may be, for example, an insurance product or a security. In addition to these, for example, it is possible to respond to various services that require examination at the time of application, such as examination of credit cards.

  1 is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read into a computer system and executed. You may perform the process of. Here, the “computer system” includes an OS and hardware such as peripheral devices.

Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Further, the “computer-readable recording medium” refers to a volatile memory (RAM) in a computer system that becomes a server or a client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. In addition, those holding programs for a certain period of time are also included. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

  The embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and includes designs and the like that do not depart from the gist of the present invention.

DESCRIPTION OF SYMBOLS 100 Source code processing apparatus 101 Source code storage part 102 Syntax analysis part 103 Nest structure analysis part 104 Priority order setting part 105 Similar product determination part 106 New product data storage part 107 Change determination part 108 Display part 109 Source code change part 110 Operation part 120 Source code 200 Abstract syntax tree 300 Nest information table 400 Priority order information 500 Existing product attribute table 600 New product attribute table

Claims (9)

A parsing unit for generating an abstract syntax tree by inputting a source code in which a determination rule for determining the suitability of application content for a product based on the attribute of the product is described for each existing product;
By analyzing a nested structure using the generated abstract syntax tree, a nested structure analysis unit that sets an order relationship between the attributes for each existing product;
A priority order setting unit that sets a priority order for each attribute described in the source code based on the set order relationship between the attributes;
Based on the priority set for each attribute and the matching state between the attribute content of the existing product and the content of the attribute of the new product, a similar product most similar to the new product is selected from the existing products. A similar product determination unit for determining,
An information processing apparatus comprising: Based on the location where the determination rule for the similar product is described in the source code, further includes a change determination unit that determines content to be changed in the source code for the addition of the determination rule for the new product.
The information processing apparatus according to claim 1. The nest structure analysis unit
Setting an order relationship between the attributes by setting a weighting value for the attribute for each existing product based on the content of the conditional expression including the attribute in the source code;
The priority order setting unit
Set a priority order for each attribute described in the source code based on a weighting value set for the attribute for each existing product,
The similar product determination unit
Determining an existing product having the highest degree of similarity determined based on the weighting value set to an attribute whose contents match between the new product and each of the existing products as the similar product,
The information processing apparatus according to claim 1, wherein the information processing apparatus is an information processing apparatus. The nest structure analysis unit
When the same attribute is included in a plurality of conditional statements in the same hierarchy, the weighting value for the same attribute is set to be greatly changed.
The information processing apparatus according to claim 3. The nest structure analysis unit
When there are a plurality of values indicating the authenticity of the attribute included in the conditional expression, the weighting value for the attribute is set to be changed to be small.
The information processing apparatus according to claim 3, wherein the information processing apparatus is an information processing apparatus. The priority order setting unit
Setting the priority order based on the transition rule about the order relation between the attributes analyzed as the nested structure;
The similar product determination unit
As a result of comparing the attribute contents until the attribute contents do not match between the new product and each of the existing products in order from the highest priority, the existing products with the largest number of attributes with matching contents Is determined as the similar product,
The information processing apparatus according to claim 1. The nest structure analysis unit
Setting an order relationship between the attributes by setting a weighting value for the attribute for each existing product based on the content of the conditional expression including the attribute in the source code;
The priority order setting unit
As a result of setting the priority order based on the transition rule, for a plurality of attributes having the same priority order, a different priority order is set using the weighting values set for the plurality of attributes.
The information processing apparatus according to claim 6. A parsing step of generating an abstract syntax tree by inputting a source code in which a determination rule for determining the suitability of the application content for a product based on the attribute of the product is described for each existing product;
Nest structure analysis step for setting the order relation between the attributes for each existing product by analyzing the nested structure using the generated abstract syntax tree;
A priority order setting step for setting a priority order for each attribute described in the source code based on the set order relation between the attributes;
Based on the priority set for each attribute and the matching state between the attribute content of the existing product and the content of the attribute of the new product, a similar product most similar to the new product is selected from the existing products. A similar product determination step for determining;
An information processing method comprising: On the computer,
A parsing step of generating an abstract syntax tree by inputting a source code in which a determination rule for determining the suitability of the application content for a product based on the attribute of the product is described for each existing product;
Nest structure analysis step for setting the order relation between the attributes for each existing product by analyzing the nested structure using the generated abstract syntax tree;
A priority order setting step for setting a priority order for each attribute described in the source code based on the set order relation between the attributes;
Based on the priority set for each attribute and the matching state between the attribute content of the existing product and the content of the attribute of the new product, a similar product most similar to the new product is selected from the existing products. A similar product determination step for determining;
A program for running

Images