TWI674511B - Product information display system, product information display method, and program product - Google Patents

Abstract

Tips have information to find the product you want. The acquisition means (101) of the product information display system (1) acquires product data indicating each category of a plurality of products that can be searched by a keyword. The specific means (102) specifies keywords for each of a plurality of users, and specifies the category of the product selected from the search results based on the product data. The recording means (103) means that if a plurality of categories are identified for the same keyword by the specific means (102), the combination of the plurality of categories is recorded in the memory means (100). The display control means (104) displays information based on the combination on the display means (25).

Description

Product information display system, product information display method, and program product

[0001] The present invention relates to a commodity information display system, a commodity information display method, and a program product.

[0002] Conventionally, in online shopping, a technique for retrieving a product according to a search condition input by a user is known. For example, Patent Document 1 describes a system that searches for a product based on a keyword input by a user, or searches for a product based on a category selected by the user. [Prior Art Document] [Patent Document] [0003] [Patent Document 1] Japanese Patent Laid-Open No. 2014-96155

(Problems to be Solved by the Invention) 0004 [0004] However, in the search based on keywords, since products are searched by using a simple string of characters, products that are not related to the desired product may be hit inadvertently. Situation. On the other hand, in the search by category, since most products belong to the category selected by the user, irrelevant products are still searched. In addition, most of the desired products are scattered in plural categories that are related to each other, rather than concentrated in one category. Therefore, even if the user selects a category, only a part of the desired product is displayed, and the products in other related categories are not displayed. [0005] At this point, even if the category names are similar, most of the types of products in them are completely different, so it is difficult for users to judge which category is related. In the technique of Patent Document 1, it is impossible to provide the user with information on the combination of the categories of relatedness. As described above, the search results are irrelevant for a large number of unrelated products, or only a part of the desired products are searched As a result, it is difficult for a user to find a desired product. [0006] The present invention has been made in view of the above problems, and an object thereof is to present information for finding a desired product. (Means for solving problems) [0007] In order to solve the above-mentioned problems, the product information display system of the present invention is characterized by including: obtaining means for obtaining product data indicating each category of a plurality of products that can be searched by keywords; The specific means is for the keywords entered by a plurality of users, and the category of the product selected from the search results is specified according to the aforementioned product data; the recording means is for the same keyword to be identified by the aforementioned specific means When a plurality of types are specified, a combination of the plurality of types is recorded in a memory means; and a display control means for displaying information based on the combination on the display means. [0008] The product information display method of the present invention is characterized in that it includes: an acquisition step of acquiring product data representing each category of a plurality of products that can be searched by a keyword; and a specific step of inputting each of the plurality of users separately. Keywords, according to the aforementioned product data to specify the category of the product selected from the search results; the recording step, if the same keyword is specified in the plural category by the aforementioned specific step, the The combination is recorded in the memory means; and a display control step for displaying the information based on the foregoing combination on the display means. [0009] The program product of the present invention causes a computer to function as a means for obtaining: obtaining means for obtaining commodity data representing each category of a plurality of goods that can be searched by a keyword; and a specific means for a plurality of users The keywords entered separately specify the category of the product selected from the search results based on the aforementioned product data; the recording means is that if plural categories are identified for the same keyword by the aforementioned specific means, the plural is classified The combination of categories is recorded in the memory means; the display control means is to display the information based on the aforementioned combination on the display means. [0010] In addition, the information storage medium of the present invention is a computer-readable information storage medium storing the above program. [0011] Further, in one aspect of the present invention, the recording means is characterized in that the attributes and attribute values of the plurality of products included in the plurality of categories included in the plurality of categories shown in the combination are extracted and It is recorded in the memory means, and the display control means causes the information based on the attribute and the attribute value to be displayed on the display means. [0012] Furthermore, in one aspect of the present invention, the aforementioned product description includes at least one of a description in a table format and a semi-structured text format description, and the aforementioned recording means is the aforementioned in the aforementioned product description At least one of the description in the form of a table and the description in the form of the semi-structured text described above, extracts the aforementioned attributes and the aforementioned attribute values. [0013] In addition, in one aspect of the present invention, the aforementioned product description further includes: a description freely inputted in the form of the text, and the aforementioned recording means is based on the aforementioned attributes and the aforementioned attribute values extracted from at least one commodity by other commodities. The aforementioned free input in the form of this text extracts the aforementioned attributes and the aforementioned attribute values. [0014] In addition, in one aspect of the present invention, the aforementioned recording means judges whether different plural plural attributes mentioned above have the same meaning as each other, and combines the plural plural attributes determined to have the same meaning. [0015] Furthermore, in one aspect of the present invention, the recording means determines whether the plural attribute has a similarity based on the number of types of common attribute values among the plurality of attributes, and whether or not the plural attribute has a threshold value or more, to determine whether the plural attribute is Have the same meaning as each other. [0016] Furthermore, in one aspect of the present invention, the display control means provides a screen for each user to select at least one of the attribute and the attribute value, and displays the information according to a selection result of the screen. The aforementioned display means. [0017] Furthermore, in one aspect of the present invention, the display control means searches for other products that match at least one of the attribute of the product and the attribute value of the product page requested by each user to display the search. Links to other products are included on the product page, so that the information is displayed on the display means. [0018] In addition, in one aspect of the present invention, the specific means or the recording means determines a category to be processed based on the number of products belonging to each category. [0019] In addition, in one aspect of the present invention, each commodity belongs to a category defined by a tree structure, and if the aforementioned recording means specifies a category of a plurality of sibling nodes for the same keyword by using the aforementioned specific means, The combination of the categories of the plural sibling nodes is recorded, and when the category of the parent node of the combination is selected, the display control means displays the information on the display means. (Effects of the Invention) [0020] According to the present invention, information for finding a desired product is presented.

[0022] [1. Hardware Configuration of Product Information Display System] An example of an embodiment of the product information display system of the present invention will be described below. FIG. 1 is a diagram showing the overall configuration of a product information display system. As shown in FIG. 1, the commodity information display system 1 includes: a server 10 and a user terminal 20, which are connected through a network so as to transmit and receive data. Among them, in FIG. 1, one server 10 and one user terminal 20 are shown, but these may be plural. [0023] The server 10 is a server computer and includes, for example, a control unit 11, a memory unit 12, and a communication unit 13. The control unit 11 includes at least one processor. The control unit 11 executes processing in accordance with a program or data stored in the memory unit 12. The memory unit 12 includes a main memory unit and an auxiliary memory unit. For example, the main memory section is a volatile memory such as a RAM, and the auxiliary memory section is a non-volatile memory such as a hard disk or a flash memory. The communication unit 13 includes a communication interface for wired communication or wireless communication, for example, to perform data communication through a network. [0024] The user terminal 20 is a computer operated by the user, such as a personal computer, a portable information terminal (including a tablet computer), or a mobile phone (including a smartphone). The user terminal 20 includes a control unit 21, a memory unit 22, a communication unit 23, an operation unit 24, and a display unit 25. The hardware configuration of the control unit 21, the memory unit 22, and the communication unit 23 may be the same as those of the control unit 11, the memory unit 12, and the communication unit 13, respectively. The operation unit 24 is an input device for a user to perform operations, such as a pointing device such as a touch panel or a mouse, or a keyboard. The operation unit 24 transmits the operation content of the user to the control unit 21. The display section 25 is, for example, a liquid crystal display section or an organic EL display section. [0025] Among them, the programs and data described as being stored in the memory sections 12 and 22 can also be supplied to these through the network. In addition, the hardware configurations of the server 10 and the user terminal 20 are not limited to the examples described above, and various types of hardware can be applied. For example, the server 10 and the user terminal 20 may also include a reading section (such as a disc drive or a memory card slot) that reads a computer-readable information storage medium, or an output for directly connecting to an external device. Access section (such as a USB terminal). At this time, the program or data stored in the information storage medium may be supplied to the server 10 or the user terminal 20 through the reading section or the input / output section. [2. Outline of Product Information Display System] The product information display system 1 can search for products processed by online shopping malls. In the product information display system 1, a search using a keyword, a search using a category, and a combination of these may be performed. For example, when the user operates the user terminal 20 and accesses the server 10, the homepage of the online shopping mall is displayed on the display unit 25, and the product can be searched. [0027] FIG. 2 is a diagram showing how a user searches for a product. As shown in Fig. 2, the top page G1 is capable of keyword input and category selection. In this embodiment, categories are defined in a tree structure, and each category is equivalent to a node. For example, only a part of categories are displayed on the first page G1. If the user selects a category to be displayed on the homepage G1, the category of the child nodes of the category can also be selected. [0028] For example, if the user selects a so-called "beer / drink" category from the home page G1, the category of "beer", "whiskey", and "wine" as a child node of the category may be selected. Then, if the user selects the so-called "wine" category, he can select the categories of "red wine", "white wine", and "wine supplies" as children of the category. After that, you can also select each category in sequence until the category of the leaf node where no child node exists. [0029] For example, when a user selects a category of "wine" and enters a keyword of "Bordeaux", as shown in the search result screen G2 of FIG. 2, a list of products for which these two conditions are searched is displayed. . Here, in terms of the sub-nodes of "wine", there are 3 "red wine", "white wine", and "wine supplies". Therefore, the search result screen G2 shows that any person belonging to these 3 categories is searched for. One of the so-called "Bordeaux" keywords. [0030] In the example of FIG. 2, a product A belonging to Bordeaux red wine, a product B belonging to Bordeaux white wine, and a product C belonging to a wine glass produced in Bordeaux are displayed on the same search result screen G2. That is, on the search result screen G2, the products C which are not wine are displayed together with the products A and B belonging to the wine itself. Bordeaux-produced wines are very popular in the world. Although the above conditions are input by most users, each user has a high probability of looking for the wine itself and a low probability of looking for wine products. Therefore, almost all of the search result screen G2 migrates to the product page G3 of product A or the product page G4 of product B, and does not migrate to the product page of product C that was inadvertently searched. [0031] The server 10 analyzes a product selected by each user from the search result screen G2, and specifies a combination of related categories. The category of relatedness refers to the category in which common items exist in product characteristics. In other words, the category of relevance is a category with similar characteristics of the product that the user is interested in. For the example in Figure 2, among the three categories of sibling nodes, the "red wine" and "white wine" often selected are the categories to which the wine itself belongs, which is a combination of related categories. On the other hand, the unselected "wine supplies" is a category that wine does not belong to, and is a category that is not related to other sibling nodes. As shown above, even if the definition of the category is in a fraternal relationship, only the part of the description is consistent, there are actually weak categories. Therefore, the server 10 is based on the user ’s actual selection of the product from the search result screen G2. The combination of categories, the relevance of specific categories. [0032] In this embodiment, the server 10 records the category of the product selected on the search result screen G2 for each keyword that has been input in the past. Therefore, the server 10 can specify, for each category, what keywords are input when a product is selected. The server 10 judges the relevance in the plural category if there is a common keyword having a threshold value or more among the plural categories. The threshold may be arbitrary, may be 1, or may be 2 or more. For example, if it is the category of "wine" in Fig. 2, if the threshold value is set to 2, for two or more common keywords, if a product of "red wine" and a product of "white wine" are selected, "red wine" The combination with "baijiu" is specified as a related category. [0033] Among them, in the above description, it is explained that even if one product of a certain category is selected, it becomes a candidate for a combination of relatedness, but if a plurality of products are selected, it can also become a relatedness combination. Waitlist. The number may be arbitrary, may be two, and may be three or more. In this way, for example, even if the user mistakenly selects the product of "wine supplies", it is possible to prevent the "wine supplies" from being included in the combination of related categories due to the wrong selection. [0034] The server 10 is based on a specific relationship between "red wine" and "white wine". In order to find the common items of the products in these two categories, refer to the product description of all products belonging to "red wine" and The description of all the products in "baijiu" extracts the attributes and attribute values of each product. The attribute is a character string that classifies the product characteristics, and the attribute value is a character string or a numerical value indicating its specific content. [0035] For example, the server 10 extracts the attribute "origin" and the attribute value "France", the attribute "production year" and the attribute value "2016", and the attribute "" from the product description of the product A (the product page G3 in FIG. 2). "Type" and attribute value "red", and attribute "degree" and attribute value "11%". In addition, for example, the server 10 extracts the attribute "origin" and the attribute value "France", the attribute "production year" and the attribute value "2015", and the attribute "type" from the product description of the product B (the product page G4 in FIG. 2). And attribute value "white", and attribute "degree" and attribute value "12%". The server 10 memorizes the extracted attributes and the combination of attribute values in association with the product, and provides the attributes and attribute values to the user. 3 is a diagram showing how attributes and attribute values are provided to a user. As shown in FIG. 3, after the homepage G1 is displayed, if the user selects the category of "wine" as a child node of "beer / alcohol", the user transitions to the attribute display screen G5, and displays the "wine""And" Liquor "and the representative attribute value of the attribute. [0037] For example, if the user selects the attribute value "France" of the attribute "origin" from the attribute display screen G5, a list of products whose attribute value is "France" is displayed on the search result screen G6. In addition, for example, if the user selects the attribute "production year" from the attribute display screen G5, a list of products for which the attribute of "production year" is extracted is displayed on the search result screen G7. Among them, at this time, since only the attribute of "production year" is selected, the attribute value is not specifically referred to. [0038] In the search result screens G6 and G7, the search results are not searched for the products whose attributes and attribute values are not extracted, so unlike the keyword search, irrelevant products can be avoided from being inadvertently searched because the text strings are consistent. Situation. For example, since the products of "wine products" which are set as unrelated are not extracted with attributes and attribute values, the products of "wine products" are not displayed on the search result screens G6 and G7. In addition, attributes and attribute values are common between "red wine" and "white wine", so you can look for goods across these categories. In addition, since attributes and attribute values are extracted from the product description, compared with the keywords of attributes prepared in advance by the system administrator, it is possible to provide more accurate information indicating product characteristics. [0039] As described above, the product information display system 1 specifies a combination of related categories based on a product selected based on a search result of a keyword input by a user. Next, the product information display system 1 provides the attributes and attribute values extracted from the combination to the user. Therefore, while preventing some irrelevant products from being displayed in the search results, it is possible to provide information across categories related to each other, and information for finding a desired product can be provided. The configuration of the product information display system 1 will be described in detail below. [3. Functions implemented in this embodiment] FIG. 4 is a functional block diagram showing an example of functions implemented in the product information display system 1. As shown in FIG. 4, in this embodiment, the data storage unit 100, the acquisition unit 101, the identification unit 102, the recording unit 103, and the display control unit 104 are implemented by the server 10. The data memory unit 100 is mainly implemented by the memory unit 12, and the acquisition unit 101, the specific unit 102, the recording unit 103, and the display control unit 104 are mainly implemented by the control unit 11. [0041] [3-1. Data Memory Unit] The data memory unit 100 stores various data about online shopping malls. Here, the category definition data, product data, search query data, PV (Property-Value) seed data, and extended product metadata are described as data stored in the data storage unit 100. 5 is a diagram showing an example of category definition data. As shown in FIG. 5, the category definition data defines master data of all categories prepared by the online shopping mall. As described above, in this embodiment, since each product belongs to a category defined by a tree structure, the category definition data defines the tree structure of the category. A tree of a category has a plurality of levels (depth). [0043] For example, the category definition data is a path for storing the category in association with the category ID of the univocally identifying category. In terms of the category path, the information from the root node to the category is stored. Among them, the character string ">>" in FIG. 5 means branching between nodes. As shown in FIG. 5, the higher the hierarchy, the more general the category name, the lower the hierarchy, and the more specific the category name is. [0044] FIG. 6 is a diagram showing an example of product information. As shown in FIG. 6, various kinds of information about a commodity are stored in the commodity data. For example, a product name stores a product name, a category ID, and a product description in association with the product ID of a uniquely identified product. Among them, the product information may store various information such as the title of the product, the store where the product is sold, the unit price, the number of stocks, or the URL of the product page. [0045] The product description is input by the store operator, and includes, for example, a description in the form of a table, a semi-structured description in the form of the text, and a description in the form of the free text. Among them, the product description does not necessarily include all of them, and may include only one or two of them. For example, a product description may include at least one of a description in a tabular form and a description in a semi-structured form. The description in a table format stores the combination of attributes and attribute values entered by the store contractor. The number of rows and columns constituting the table can be arbitrary. The description in the semi-structured text is the text in which the store contractor enters a part of the stereotype. For example, prepare plural and semi-structured texts, and decide in advance the position of words that a store contractor can enter. The description in the form of free input text is a text freely input by the store store operator. You can also enter HTML tag information for free entry in this article. [0046] FIG. 7 is a diagram showing an example of search query data. As shown in FIG. 7, the search query data stores the relationship between the search conditions entered by the user and the actions of the user on the search results. For example, in the search query data, the keywords entered by the user are related to the product ID of the product actually selected by the user from the search result, and the category ID of the product. [0047] Among them, the keyword search can also combine plural keywords, so in the search query data, the conditions of each of the plural keywords and the combination of these keywords ("AND" or "OR") can be stored. In addition, not only keywords but also categories selected by the user when searching can be stored in the search query data. [0048] In this embodiment, each time a user enters a keyword to perform a search, the search query data is updated. For example, if the server 10 receives a keyword input by a user, the server 10 stores the keyword in the search query data. Next, the server 10 receives the selection result of the search result, and stores the product ID and the category ID of the product selected by the user in the search query data. 8 is a diagram showing an example of PV seed data. PV seed refers to the master data that defines a combination of attributes and attribute values. That is, the PV seed system functions as a dictionary (sample) that is referred to in order to extract a combination of attributes and attribute values from the product description. For example, a combination of a PV seed profile with a category of relatedness results in a related storage of PV seeds extracted from that combination. [0050] However, in this embodiment, even if they have the same meaning, different attributes are integrated into one. For example, the attribute of "origin" may be described as "origin" or "origin" depending on the store. These meanings are the same, so they are integrated into one attribute as shown in FIG. 8. Therefore, a certain attribute may be composed of plural types of notations. This point system has the same attribute value, and one attribute value can also be composed of multiple types of notation. [0051] In addition, information identifying representative representations among the integrated attributes may be stored in the PV seed data. For example, among "origin", "origin", and "origin", "origin" can be set as a representative token. In addition, since a plurality of attribute values exist in a certain attribute, information identifying a representative attribute value among them can be stored in PV seed data. For example, among the attribute "origin", "France", "Italy", "Spain", etc., which appear mostly as attribute values, can also be set as representative attribute values. The representative attributes and attribute values can be determined according to the number of these extracted products, and can also be designated by the system administrator. If it is determined based on the number of products, the attribute with the largest number of products and the attribute value may become representative values, or according to the order of increasing number of products, the predetermined number may become the representative value. [0052] FIG. 9 is a diagram showing an example of extended commodity metadata. As shown in FIG. 9, the extended commodity metadata is data indicating attributes and combinations of attribute values associated with each commodity. For example, when expanding product metadata, a combination of attributes and attribute values is stored in association with the product ID. The attributes and attribute values stored in the extended product metadata are among the attributes and attribute values stored in the PV seed data, and are found by the commodity description of each commodity. [0053] However, the data stored in the data storage unit 100 is not limited to the example described above. The data storage unit 100 is only required to memorize data required for processing in the implementation mode. For example, the data storage unit 100 may store HTML data of each screen. [3-2. Acquisition Section] The acquisition section 101 acquires product data indicating each category of a plurality of products that can be searched by a keyword. The acquisition unit 101 acquires product data from the data storage unit 100. When the data storage unit 100 is located in a system different from the product information display system 1, the obtaining unit 101 can obtain the product data from the different system. [3-3. Specific section] The specific section 102 is a keyword (a keyword may be one or a plurality of keywords) input by a plurality of users, and is specified by a search result based on product data. The category of the selected product. In this embodiment, since the category ID is stored in the search query data updated based on the product data, the specifying unit 102 refers to the search query data and specifies the category of the product actually selected by each user who entered the keyword from the search result. . If the category ID is not stored in the search query data, the specific unit 102 refers to the product data to obtain the category ID associated with the product ID stored in the search query data. [3-4. Recording Unit] If the recording unit 103 specifies a plural category for the same keyword (the keyword may be one or plural) by the specific unit 102, the The combination is recorded in the data memory section 100. In other words, the recording unit 103 specifies a plural category as a combination of relatedness when there is a common keyword in the plural category. [0057] In the present embodiment, the recording unit 103 generates, based on a combination of categories, search information capable of performing a search across a plurality of categories included in the combination. As long as the search information is information that can be used in product search, it may be a character string, a numerical value, or a combination of these. Retrieval information is searched across plural categories of relatedness, and products in unrelated categories are not found. For example, the search information generated by the products in the plurality of related categories is related to the products in the plurality of categories, and is not related to the products in the other categories. The search information can be described as a common item of products in a combination of related categories. The search information may not be a combination of attributes and attribute values, a character string indicating the characteristics of a certain product, or a character string that is part of a product name. [0058] In addition, a combination of categories related to connectivity may be used for information presentation, and may not be used for retrieval. For example, the combination of categories recorded by the recording unit 103 may be directly displayed on the display unit 25, and the extracted attributes and attribute combinations may be presented as information only in a non-retrievable state. In addition, a combination of related-type categories may be presented as information together with keywords that are referred to when specifying the combination. [0059] In this embodiment, a case where the recording unit 103 mainly executes the following six processes will be described. (1) Category extraction processing (2) Category integration processing (3) PV seed extraction processing (4) Attribute integration processing (5) Attribute value expansion processing (6) Link processing [0060] [Class extraction processing] The category extraction processing is specific The process of the target category after the category integration process. That is, the category extraction process is a process of specifying a category that is a target for determining relevance. In other words, the category extraction processing is processing that specifies a category that is an object of a combination of an attribute and an attribute value. [0061] For example, if there are too many products in the category, because the products are too diverse, even if the attributes and attribute values are extracted, it is not very meaningful. On the contrary, if the number of products is too small, because the product categories are too small, even if the attributes and Attribute values are also less helpful. Therefore, in this embodiment, a category with an appropriate number of products is used as a processing target by the category extraction processing. [0062] For example, the recording unit 103 is based on the category definition data and the product data, and among the entire tree, the number of specific products is the first threshold (for example, 2 ⁷ ) Above and less than the 2nd threshold (e.g. 2 ¹⁵ ) Part of the tree, and extract the type of the root node of the part of the tree. Among them, the recording unit 103 is not only the root node, but also all types included in the part of the tree. In addition, the first threshold value and the second threshold value are stored in the data storage unit 100 and may be fixed values or variable values. [Category Integration Process] The category integration process is a process of specifying a combination of related categories. That is, the category integration process is a process of combining a combination of categories related to connectivity into one group. The recording unit 103 performs category integration processing based on the categories extracted in the category extraction processing and the search query data. [0064] In this embodiment, the category extraction processing is used to extract the roots of a part of the wood to which a certain number of products belong. Therefore, in the category integration processing, the recording unit 103 totals the keywords entered when the products of the category are selected for each category of the child node of the root node. Next, the recording unit 103 specifies the plural categories as a combination of related categories if there is a reference number (for example, 30) or more of the keywords common to the plural categories. The recording unit 103 stores the combination of the identified categories in the PV seed data. [0065] However, the method for determining the correlation between categories is not limited to the above-mentioned example. For example, if the recording unit 103 specifies a plurality of categories for a certain keyword and the number of users who have entered the keyword is equal to or greater than the reference number, it may be determined that there is a correlation between the categories. That is, for example, instead of making selections for only a certain group of users, it is also possible to set the categories of products selected by most users as a combination of related categories. In addition, as described above, when a plurality of products in a category are selected for a certain keyword, the category can also be used as a candidate for determination of relevance. [PV Seed Extraction Process] The PV seed extraction process is a process of extracting PV seeds from the product description. The PV seed extracted by the PV seed extraction process is an initial value, and the attribute integration process and the attribute value expansion process described later are used to integrate contractual attributes or increase changes in attribute values. [0067] The recording unit 103 executes PV seed extraction processing based on a combination of the categories specified in the category integration processing and the product data. For example, in the algorithm for extracting a PV seed, a place where attributes are stored in a product description is set. The storage location of the attribute may be set in advance based on the tendency of the product description input in the past. The recording unit 103 extracts a character string of a place set by the algorithm as an attribute in the product description. In addition, in the algorithm, a relative storage location of an attribute value versus an attribute storage location is set. The recording unit 103 extracts, as the attribute value, a character string or a numeric value located in a place having a predetermined positional relationship with the extracted attribute. [0068] For example, the description in the form of a table is a character string inputted for each cell. Therefore, compared with the product description in the form described herein, it is easier to specify the storage place of the attribute and the attribute value. Therefore, the recording unit 103 can extract attributes and attribute values from the description in the form of a table. In the description in the form of a table, an attribute is mostly stored in a specific column (for example, the first column), and an attribute value is stored in an adjacent column (for example, the second column). Therefore, the recording unit 103 obtains the character string stored in the cell of the specific row as an attribute, and obtains the character string stored in the adjacent row as the attribute value. Next, the recording unit 103 extracts these combinations as PV seeds and stores them in the PV seed data. [0069] For example, in the case of the products A and B of FIG. 2, as shown in the product pages P3 and P4, in the product description in the form of a table, the attribute name is stored in the first column and the attribute value is stored in the second column. Therefore, the recording unit 103 obtains the character string of each cell stored in the first row as an attribute, and obtains the character string of each cell stored in the second row as an attribute value. For example, the recording unit 103 extracts the attribute "origin" and the attribute value "France", the attribute "production year" and the attribute value "2016", the attribute "kind" and the attribute value "red", and the attribute "degree" from the product A. The four combinations of "" and the attribute value "11%" are used as PV seeds. In addition, for example, from the product B, the attribute "origin" and the attribute value "France", the attribute "production year" and the attribute value "2015", the attribute "kind" and the attribute value "white", and the attribute "degree" and the attribute are extracted Four combinations with a value of "12%" are used as PV seeds. [0070] Among them, depending on the table, there are cases where there is no specific column and attributes are stored in a specific row. Therefore, at this time, the recording unit 103 obtains the character string of the cell stored in a specific row as an attribute and will be The character string of the cell stored in the adjacent row is obtained as the attribute value. [0071] In addition, for example, the semi-structured explanation in the form of the text is fixed in the place where the word is inserted, so it is easy to specify the storage place of the attribute and the attribute value. Therefore, the recording unit 103 can also extract PV seeds based on the semi-structured text. The semi-structured text has a fixed overall format, so the word inserted at the first position is an attribute, and the word inserted at the second position is a high probability. Therefore, the recording unit 103 may also obtain the attribute by referring to the first position of the semi-structured text, and obtain the attribute value by referring to the second position, and extract the combination as a PV seed. [0072] In addition, the recording unit 103 may expand the extracted PV seed by using a description freely input in the form of this text. At this time, the recording unit 103 determines whether there is a character string similar to at least one of the attributes and attribute values of the PV seeds extracted from the description in the form of a table or the semi-structured text in the description freely entered in the text. . For example, if the character string similar to the current PV seed is included in the description of the article format, the recording unit 103 adds the character string as a new attribute to the PV seed. In addition, for example, if the character string or numerical value similar to the attribute value of the current PV seed is in the description of the article, the recording unit 103 adds the character string or numerical value to the PV seed as a new attribute value. [0073] Among them, the method for determining the similarity of the character string itself is applicable to various known similarity determination algorithms. In other processes that follow, it is the same for determining whether the character string is similar or not. For example, a so-called Word2vec algorithm that returns the feature vector of the input character string may be used. For example, if the difference between the feature quantity vectors of the two character strings is less than the threshold, the judgment is similar, and if the difference is more than the threshold, the judgment is not similar. In addition, in a language like Japanese, where there is no space between words, the recording unit 103 can also use an algorithm to determine the division position of the word, divide the description of the article form into words, and execute the stored in Whether the attributes of the PV seed are similar to the attribute value is determined. In addition, the set of divided words can be used to determine the similarity of the character string, and the connective made of plural words can also be used to determine the similarity of the character string. [0074] [Attribute integration processing] The attribute integration processing is a processing that integrates the attributes of the same meaning among the attributes extracted as the PV seeds. Through attribute integration processing, the same attribute is integrated into one, and the redundancy of PV seeds is reduced. [0075] For example, even if the attributes are the same, there are different expressions. Therefore, in the attribute integration processing, plural attributes with different expressions are integrated into the same attribute. For example, if the attribute described as "manufacturing source" and the attribute described as "manufacturer" exist in PV seeds, these are only different from Japanese and English. Since they have the same meaning, they are processed by attribute integration. These are integrated into the same attributes. [0076] For example, in languages such as Japanese, where plural types of characters are mixed, the types of characters used may be different even for the same words. For example, if Japanese is a mixture of three types of characters: Hiragana, Katakana, and Kanji, when a word is recorded in Hiragana, when it is recorded in Katakana, and when it is recorded in Kanji, the meanings are the same, so the attributes are unified Processing, these are integrated into the same attributes. [0077] The recording unit 103 performs attribute integration processing based on the PV seeds extracted in the PV seed extraction processing. In this embodiment, a case where the similarity of attributes is determined based on the commonality of attribute values will be described. For example, the recording unit 103 calculates the similarity L between the attributes stored in the PV seed based on the following Equation 1. [0078] [0079] P of Formula 1 ₁ With P ₂ Arbitrary attributes extracted from the PV seed. That is, it is a combination of two attributes that is a target for determining whether or not to integrate attributes. The left side of Equation 1 is 2 attributes P ₁ With P ₂ Similarity L. Here, the higher the similarity L, the attribute P ₁ With P ₂ The more similar, it means that these should be integrated into the same attribute. On the other hand, the lower the similarity L, the attribute P ₁ With P ₂ The less similar, it means that these should not be integrated into the same attribute. M of Formula 1 ₁ Attribute P ₁ Range size, m ₂ Attribute P ₂ Range size. Among them, the range size refers to the range covered by the attribute, and here means the diversity of the attribute value. For example, the more types of attribute values of an attribute, the larger the size of the range, the fewer types of attribute values, and the smaller the size of the range. In addition, for example, if the attribute value is represented by a numerical value, the larger the difference between the maximum value and the minimum value of the attribute value, the larger the range size, the smaller the difference, and the smaller the range size. N system m of formula 1 ₁ With m ₂ The dimensions of the common parts. That is, for property P ₁ Attribute value and attribute P ₂ The number of types of attribute values that are common among the attribute values. L of Formula 1 _conf The system returns two functions that increase the accuracy of similar attributes. The accuracy is attribute P ₁ Attribute value and attribute P ₂ The larger the common part of the attribute value (that is, the greater the number of common attribute values), the larger, and the smaller the common part, that is, the smaller. Function L _conf System is used in valuation in attribute P ₁ With attribute P ₂ To what extent are attribute values common. L of Formula 1 _size Used to slow L _conf Shows the reduction function of the accuracy. Function L _size Attribute if P ₁ Range size, and attribute P ₂ The difference in the size of the range is small, that is, becomes large, and if the difference is large, it becomes small. Assume that even in attribute P ₁ With attribute P ₂ Among them, there are many common parts. When there is a difference in the size, it will not be determined whether they are regarded as the same. Therefore, the function L _size System is used to prevent such integration. L of Formula 1 _error If the first term on the right is a very large or very small value, it is used to adjust the increase function of the similarity L. [0083] As shown in Equation 1, the function L _conf Tie in m ₁ , M ₂ , And n are variables, and the function L _size Tie in m ₁ And m ₂ Is the formula for the variable, the function L _size It is a formula with n as a variable. For example L _conf (x) = x, L _size = exp (-ax), and L _error (x) = bx. Among them, a and b are coefficients, for example, a = 0.33, and b = 0.1. [0084] When the similarity L of two attributes is equal to or greater than a reference value (for example, 0.1), the recording unit 103 integrates the two attributes into the same type. Among them, among the categories of relatedness, the same attribute is unified, so the categories of the products from which the two attributes are extracted are related to each other, which is also a condition for integration. [0085] Among them, the attribute integration processing is not limited to the above-mentioned example, and other attributes may be used to estimate the same attribute. For example, a well-known class determination algorithm may be used, and the recording unit 103 integrates attributes determined to be synonymous based on the class determination algorithm. In the category determination algorithm, the category can also be determined based on the category dictionary data that defines the category. In addition, for example, the recording unit 103 may refer to the attribute value of each of the plurality of attributes extracted as the PV seed, and estimate attributes having specific attribute values to be the same type. The specific attribute value may be a pre-designated attribute value, but it may also be an industry term commonly used among industry players, for example. [0086] In addition, for example, it is not easy for the recording unit 103 to coexist plural expressions in one product description with respect to a certain attribute. Therefore, attributes that appear repeatedly in a product description may be limited to not be presumed to be the same type. In addition, in the attribute integration processing, the recording unit 103 may also perform a filtering process of the attribute value. In the attribute value filtering process, the attribute value that can be extracted as PV seed data is discarded, or an attribute value is integrated with other attribute values into the same kind. For example, the recording unit 103 may discard attribute values that have not appeared as keywords stored in the search query data. [0081] [Attribute Value Expansion Process] The attribute value expansion process is a process that increases the change in the attribute value and expands the range size of the attribute. That is, the attribute value expansion process is a process of adding a new attribute value to each attribute. [0088] For example, the recording unit 103 performs attribute value expansion processing based on the product data. The recording unit 103 extracts a candidate for a new attribute value from the product description of the product to which the category included in the specific combination is included in the category integration process. For example, the recording unit 103 may extract a character string or numerical value similar to the attribute value stored in the current PV seed as a candidate for the attribute value. In addition, for example, the recording unit 103 may extract a character string or a numerical value near a character string corresponding to the attribute stored in the current PV seed, as a candidate for the attribute value. [0089] The recording unit 103 determines whether or not the candidate extracted as described above is added to the PV seed. The recording unit 103 repeats this process until there are no new candidates. For example, if the recording unit 103 satisfies two conditions, the extracted candidate may be added to the PV seed as a new attribute value. The first condition is an algorithm for determining whether the candidate is similar to the current attribute value based on the similarity of the characters. The second condition is that there is no other attribute having an attribute value similar to the candidate. If there is another attribute having an attribute value similar to the candidate, the attribute value may be added to the other attribute. [0090] For example, the recording unit 103 may calculate an evaluation value based on Equation 2. The left side of Equation 2 is the evaluation value. S _P 系 Attribute. Mi is a similar algorithm for determining whether text is used. On the right, the feature vector of the characters shown in the characters and the deviation from the feature vector of the existing attribute value. Since the cosines of these are obtained as shown on the right, the higher the vectors are in the same direction (ie, the more similar they are as text), the higher the evaluation value. The recording unit 103 adds a candidate as a new attribute value if the evaluation value is greater than the threshold value, and discards the candidate whose evaluation value does not reach the threshold value. [0091] [0092] [Linking Process] The linking process is a process of assigning attributes and attribute values to products of each category set as related. As mentioned above, PV seeds are only used as dictionaries. Therefore, at the time when PV seeds are extracted, attributes and attribute values have not yet been assigned to products. Through the linking process, attributes and attribute values are assigned to the product, and expanded product metadata is generated. [0093] For example, the recording unit 103 performs link processing based on the product data and the PV seed data. The recording unit 103 extracts attributes and attribute values based on the description of each product in a table format and PV seed data. That is, in the description of each product in the form of a table, the recording unit 103 extracts attributes and attribute values that are the same as the combination of the attributes and attribute values shown in the PV seed. [0094] In addition, for example, the recording unit 103 extracts attributes and attribute values based on the semi-structured text form of each product or freely enters a description of the text form and PV seed data. That is, the recording unit 103 is among the descriptions in this text form, and if there is a match with the combination of the attribute and the attribute value shown in the PV seed, the attribute and the attribute value are extracted as the commodity. [0095] Among them, when the attributes and attribute values are extracted, the aforementioned similar determination algorithm used to determine the character string rather than the same as described above may also be used. In addition, for example, if the attribute value is a numerical value, the recording unit 103 may consider the attribute value to be in the product description if the difference between the value of the product description and the attribute value in the PV seed is less than the threshold. [0096] As described above, the recording unit 103 of the present embodiment extracts the attributes and attribute values of the products and records them in the data storage unit 100 based on the product description of each of the plurality of products included in the plurality of categories shown in combination. In this embodiment, the recording unit 103 extracts attributes and attribute values for all products in the plural category that are specified as related persons. However, only a part of the products in the plurality of categories may be formed as extraction targets of attributes and attribute values, and it is not necessary to necessarily form all the products as extraction targets of attributes and attribute values. [0097] For example, the recording unit 103 extracts attributes and attribute values from at least one of a description in a table format and a semi-structured description in a product description. For example, the recording unit 103 extracts attributes and attribute values from a description in a table format among product descriptions. The recording unit 103 extracts a character string of a specific cell in the description in a table format as an attribute, and extracts a character string or a value of another cell in a predetermined positional relationship with the cell as an attribute value. In addition, for example, the recording unit 103 extracts a character string at a predetermined position in a semi-structured description in the form of a text description as an attribute, and extracts a character string or a value at another position in a predetermined positional relationship with the position as an attribute value. This positional relationship can vary depending on the language used. For example, if the subject is an attribute and the predicate is of high probability, the attribute can be set to the initial noun of the article and the attribute value can be set to the noun before the predicate. If it is English, the attribute can be set to the initial of the article. Noun, the attribute value is set to the noun after the predicate. [0098] For example, the recording unit 103 may extract the attributes of the PV seeds from the description in a table format, and extract the attribute values of the PV seeds from both the description in the table format and the semi-structured description in the text format. The recording unit 103 may also freely enter a description in the form of a text based on a PV seed including such extracted attributes and attribute values, extract attributes and attribute values, and store them in association with the product. [0099] In addition, for example, the recording unit 103 extracts attributes and attribute values based on the attributes and attribute values extracted from at least one product, and freely enters a description in the form of this text by other commodities. The recording unit 103 may also extract attributes and attribute values of other products based on the attributes and attribute values extracted from the individual products. That is, the recording unit 103 may extract attributes and attribute values from the product Z based on the attributes extracted from the product X and the attribute values extracted from the product Y. For example, the recording unit 103 extracts text from a description that is freely input by other products in the form of text that is consistent with or similar to the attributes extracted from the tabular or semi-structured text of the description of at least one product. As a new property. Similarly, the recording unit 103 is a text string or numerical value that is consistent with or similar to the attribute value extracted from the tabular or semi-structured text description of at least one product in the description of the free input text form of other products. , It is extracted as the new attribute value. [0100] For example, the recording unit 103 determines whether or not the plural plural attributes having different meanings have the same meaning, and combines the plural plural attributes determined to have the same meaning. As described above, in the present embodiment, the recording unit 103 determines whether or not the degree of similarity L (for example, Equation 1) based on the number of types of common attribute values among plural attributes is greater than or equal to a threshold. Whether plural attributes have the same meaning as each other. For example, if the similarity L between the attributes is greater than or equal to a threshold, the recording unit 103 determines the attributes with the same meaning and integrates them. However, the method for calculating the similarity is not limited to Equation 1, and other equations may be used. The number formula may be a number formula in which the number of kinds of common attribute values is used as an argument, or may be a number with a higher number of kinds. Alternatively, for example, the number of types of common attribute values may be directly similarity. In addition, other methods may be used to determine similar words. For example, the category dictionary data may be stored in the data storage unit 100, and the recording unit 103 integrates plural attributes based on the category dictionary data. [0101] For example, the recording unit 103 determines a category to be processed based on the number of products belonging to each category. The recording unit 103 determines the relationship with other categories, or extracts attributes and attribute values for categories with a certain number of products. In addition, for example, in this embodiment, since the category has a tree structure, the recording unit 103 records the combination of the categories of the plural sibling nodes when the category of the plural sibling nodes is specified by the specific unit 102 for the same keyword. . Next, the recording unit 103 generates a combination of an attribute and an attribute value when a combination of categories of sibling nodes is specified by the specifying unit 102 for the same keyword. [3-5. Display Control Unit] The display control unit 104 displays the information based on the combination on the display unit 25. For example, the display control unit 104 displays information based on attributes and attribute values on the display unit 25. For example, the display control unit 104 displays the search result based on the person selected by the user among the attributes and attribute values displayed on the display unit 25. In addition, for example, as described above, the display control unit 104 may display the extracted attributes and attribute values on the display unit 25, and may display information other than the retrieved information on the display unit 25. Alternatively, for example, the display control unit 104 may be included in the category tree, and the related sub-categories may be displayed on the display unit 25 in a recognizable manner. The information provided by the display control unit 104 may also be the manager of the online shopping mall, not only the user of the online shopping mall. That is, the display control unit 104 may display information on an arbitrary display unit other than the display unit 25 of the user terminal 20. [0103] In the present embodiment, the display control unit 104 performs a search for each category shown by a combination of categories of relatedness based on a combination of attributes and attribute values. A cross-category search refers to a cross-type search of a plurality of categories, and simultaneously searches for products in the first category and products in the second category based on a certain search condition. [0104] For example, the display control unit 104 provides attribute display screens G5 and G6 of at least one of selectable attributes and attribute values to each user, and displays a search result based on a selection result of the screen on the display unit 25. . For example, the display control unit 104 may perform a search across the categories shown in the combination based on the selection results of the attribute display screens G5 and G6. When the user selects an attribute, the display control unit 104 refers to the expanded product metadata, and displays the product related to the attribute in the search result. In addition, when the user selects an attribute value, the display control unit 104 refers to the expanded product metadata, and displays the product related to the attribute value in the search result. [0105] In the present embodiment, since categories are defined in a tree structure, the display control unit 104 displays information on the display unit 25 when a category of a parent node of a combination of categories related to connectivity is selected. For example, when the category of the parent node is selected by the display control unit 104, based on the combination of the attribute and the attribute value, the search is performed for each category shown by the combination of the categories of relatedness. That is, the display control unit 104 displays the attribute display screens G5 and G6 showing the attributes and attribute values extracted from the category related to the child nodes of the category selected by the user. As a result of the selection of the screen, a search is performed across the child nodes of relatedness. [0106] [4. Processing executed in this embodiment] Next, a processing flow executed in the product information display system 1 will be described. Here, a description will be given of a metadata generation process for generating an expanded product metadata by identifying a combination of related categories, and an information display process for searching for a product based on the expanded product metadata. Metadata generation processing and information display processing are examples of processing performed by the functional blocks shown in FIG. 4. [4-1. Metadata Generation Processing] FIG. 10 is a flowchart showing metadata generation processing. The metadata generation process is executed by the control unit 11 in accordance with a program stored in the memory unit 12. The metadata generation processing may be performed at a predetermined timing. For example, the metadata generation processing may be performed according to an instruction of a system administrator, or may be performed according to a preset date and time. [0108] As shown in FIG. 10, first, the control unit 11 executes a category extraction process based on the category definition data and the product data (S1). In S1, the control unit 11 specifies a category having a certain number of products from all the categories, and extracts the category as a subsequent processing target. The control unit 11 executes a category integration process based on the category extracted in S1 and the search query data (S2). In S2, the control unit 11 specifies a combination of the categories of the products on which the product page is displayed using a common keyword, and holds them as a combination of related categories. [0109] The control unit 11 executes a PV seed extraction process based on the combination of the categories specified in the category integration process and the product data (S3). In S3, the control unit 11 extracts the PV seed from the product description in the form of a table or a semi-structured text, and expands it in the product description freely entered in the text. The control unit 11 executes attribute integration processing based on the PV seed data (S4). In S4, the control unit 11 determines whether or not the plural complex attributes having different meanings have the same meaning, and integrates plural complex attributes with the same meaning. [0110] The control unit 11 executes attribute value expansion processing based on the product data (S5). In S5, the control unit 11 extracts a new attribute value from the product description and adds it to the PV seed. The control unit 11 executes a connection process (S6) based on the product data and the PV seed data, and this process ends. In S6, the control unit 11 extracts the combination of the attribute and the attribute value, and records it in the expanded commodity metadata in association with the product if there is a product that is identical or similar to the PV seed in the product description. [4-2. Information Display Processing] FIG. 11 is a flowchart showing the information display processing. The information display processing is performed by the control unit 11 in accordance with a program stored in the storage unit 12, and the control unit 21 operates in accordance with a program stored in the storage unit 22. [0112] As shown in FIG. 11, first of all, in the user terminal 20, the control unit 21 sends a display request of the home page G1 to the server 10 in accordance with the user's selection of the home page G1 by the operation unit 24 (S10 ). Among them, if the user terminal 20 sends data to the server 10, the IP address or session ID of the user terminal 20 is transmitted, and the server 10 can specify which user terminal 20 Accessing. [0113] When the server 10 receives the display request of the home page G1, the control unit 11 generates display data of the home page G1 and sends it to the user terminal 20 (S11). In S11, the control unit 11 generates a home page G1 including a selectable category based on the category definition data, or selects an arbitrary product based on the product data to generate a home page G1 including a link to the product. The homepage G1 can be any data format, for example, it can be HTML data. [0114] When the user terminal 20 receives the display data of the home page G1, the control unit 21 displays the home page G1 (S12). The control unit 21 sends the operation content of the user on the home page G1 to the server 10 based on the detection signal of the operation unit 24 (S13). Here, it is formed to perform either of input of a keyword and selection of a category, but other operations may be performed. [0115] In the server 10, when receiving the operation content of the user, the control unit 11 determines whether the user has selected a category (S14). If it is not determined that the user has selected a category (S14; N), this process ends, and the control unit 11 executes a process based on the operation content. For example, if the user has entered a keyword, the control unit 11 executes a keyword search and sends the search result to the user terminal 20. After that, when the server 10 receives information indicating the product selected by the user, the server 10 stores the keywords entered by the user in the search query data in association with the category of the product selected by the search result. [0116] On the other hand, when it is determined that the user has selected a category (S14; Y), the control unit 11 generates display data of the attribute display screen of selectable attributes and attribute values based on the PV seed data and sends it to The user terminal 20 (S15). In S15, the control unit 11 specifies the combination extracted as the related category among the child nodes of the category selected by the user. Next, the control unit 11 generates an attribute display screen for selecting a PV seed associated with the combination. Among them, in the attribute display screen, it is not necessary to display all the PV seeds, and only representative attributes and attribute values may be displayed. [0117] When the user terminal 20 receives the display data of the screen, the control unit 21 displays the attribute display screen (S16). The control unit 21 sends the operation content of the user to the server 10 based on the detection signal of the operation unit 24 (S17). Here, it is formed as a selector for selecting an attribute or a value of an attribute, but other operations can also be performed. [0118] When the server 10 receives the operation content of the user, the control unit 11 refers to the operation content (S18). If it is determined that the user has selected an attribute (S18; attribute), the control unit 11 searches the attribute selected by the user to generate a related product based on the expanded product metadata (S19). In S19, the control unit 11 compares the attribute selected by the user with the attribute stored in the extended product metadata, and extracts the product for which the attribute selected by the user is searched. [0119] On the other hand, if it is determined that the user has selected an attribute value (S18; attribute value), the control unit 11 retrieves the attribute value selected by the user to generate a related product based on the expanded product metadata (S20) ). In S20, the control unit 11 compares the attribute value selected by the user with the attribute value stored in the extended product metadata, and extracts a product for which the attribute value selected by the user is searched. [0120] Based on the search result of S19 or S20, the control unit 11 generates display data of the search result screen and sends it to the user terminal 20 (S21). The search result screen includes thumbnail images or links of the products searched in S19 or S20. [0121] When the user terminal 20 receives the display data of the search result screen, the control unit 21 displays the search result screen (S22), and this process ends. After that, the user selects a desired product from the search result screen and transitions to the product page. [0122] With the product information display system 1 described above, a combination of related categories is recorded and information based on the combination is displayed. Therefore, even if the products are scattered in a plurality of categories, information for finding a desired product can be presented. For example, if a search across multiple categories is performed based on common search information in a combination of related categories, the products scattered in the multiple categories can be listed in one search result, and a search result that easily finds the desired product can be presented. . In addition, for example, even if the combination of the related categories is displayed as it is, the user can grasp the category in which the desired product is dispersed, and thus can present useful information for the user. In addition, the product information display system 1 is based on the relationship between specific categories based on the product actually selected by the user, so the relationship between the categories according to the actual status can be found. In addition, as described above, the specific unit 102 specifies the plural categories as a combination of relatedness if there are keywords common to the plural categories above the threshold, but if the threshold is set to In the case of 2 or more (plurality), it is possible to prevent a situation in which a non-relevant category is specified as a related person due to a user's wrong selection of a product. In addition, since it is a candidate of a related category, the condition that two or more products are selected as a condition can also prevent a situation where a non-related category is specified as a related person due to a user's mistaken selection. [0123] In addition, the product information display system 1 prompts information based on the attributes and attribute values extracted from the product description, and therefore, common items among categories related to connectivity can be presented as information. Based on the common terms, users can find hints for products scattered in plural categories. For example, if the attribute and the attribute value are used as the search information, the search may be performed based on the common terms between the related categories. Therefore, even if the related products are scattered among a plurality of categories, a desired product located in each category can be pinpointed and posted on the search result. In addition, attributes and attribute values are extracted from the product description. Therefore, compared with the keyword of the attribute prepared in advance by the system administrator, it is possible to present information that more accurately represents the characteristics of the product. [0124] In addition, if the product information display system 1 extracts the attributes and attribute values of each product from the description in a tabular format where attributes and attribute values are easily stored, it is possible to prevent information that does not indicate the characteristics of the product from being extracted by mistake. Situations to improve the accuracy of attributes and attribute values. In addition, if the product information display system 1 extracts the attributes and attribute values of each product from the semi-structured textual description of the storage place where attributes and attribute values are easily specified, it can also prevent information that does not indicate the characteristics of the product from being extracted by mistake Can improve the accuracy of attributes and attribute values. [0125] In addition, the product information display system 1 uses the attributes and attribute values extracted from at least one of the description in the tabular form and the semi-structured text form, and freely enters the attributes and attribute values in the text. Enter more information in this article to increase the change of attributes and attribute values. [0126] In addition, the product information display system 1 can prevent attributes of the same meaning from being extracted as different attributes by unifying attributes with different meanings, and can reduce attribute redundancy. [0127] In addition, the commodity information display system 1 improves the determination accuracy of synonyms by referring to the similarity of the number of types of attribute values that are common among the attributes to determine whether the attributes have the same meaning. For example, if the synonym is determined by the number of attribute values, if a store repeatedly uses the same attribute value in the product description, the number will increase. There will be a property that is not a synonym because of the product description of the store. The possibility of being made synonymous with each other. In contrast, as explained in the embodiment, if a synonym is determined by the number of types of attribute values, even if it is used several times as a product description, the number of types does not change to one, so It is possible to prevent the non-synonymous attributes from being formed as synonymous with each other as described above, and it is possible to improve the accuracy of determining synonyms. [0128] In addition, the product information display system 1 displays the attribute display screen of at least one of the selectable attribute and the attribute value on the user terminal 20, so that the user can grasp which viewpoint has common items. For example, if the search is performed based on the attribute or attribute value selected by the user, the search may be performed in accordance with the viewpoint of the user preference. [0129] In addition, the product information display system 1 determines the correlation between the categories of the child nodes corresponding to the categories with a certain number of products to extract attributes and attribute values, thereby providing information more efficiently. For example, like "drinks", when there is a general-named category, because there are too many products belonging to that category, the characteristics of the products in that category are too different, and it is meaningless to analyze attributes and attribute values. Conversely, in a category with too few products, since the desired product is found immediately, even if it is analyzed, it is meaningless. Therefore, it is possible to perform a search within a valid category by searching using attributes and attribute values. Therefore, for example, as "wine", it is an appropriate and specific category, and the category with the appropriate number of products is the processing object. In the category corresponding to the child node of "wine" (here, "red wine", "white wine") "", "Wine supplies"), through the combination of specific related categories, can provide more useful information for users. [0130] In addition, the product information display system 1 can provide information across these categories if the categories of the child nodes of the category selected by the user are related to each other. Therefore, if the user is allowed to select a rough category, the information across the related categories can be provided from the perspective of attributes and attribute values. [0131] [Modifications] However, the present invention is not limited to the embodiments described above. Appropriate changes can be made without departing from the gist of the present invention. [0132] (1) For example, in the embodiment, a description is given of a case where an attribute display screen for selecting extracted attributes and attribute values is displayed. However, a search result based on the extracted attributes and attribute values is presented to The user only needs to recommend the products retrieved based on the attribute and the attribute value to the user. [0133] The display control unit 104 of the modified example (1) searches for other products that match at least one of the attributes and attribute values of the products requested by each user to display the product page, and the product page includes other searched items. By linking the products, the information is displayed on the display unit 25. The product page request can be made by selecting a product from the search result screen. [0134] For example, the display control unit 104 may also obtain the attributes and attribute values of the products requested to display the product page based on the expanded product metadata, and search for products having the same attributes and attribute values as those, and make the products The link is included on the product page. In addition, for example, if the display control unit 104 specifies the category ID of the product requested to display the product page based on the product data, and the category ID is registered in the PV seed data, it obtains any PV seed associated with the category ID. . Next, the display control unit 104 may search for a product based on the expanded product metadata and the obtained PV seed, and include a link to the product on the product page. [0135] According to the above-mentioned modification (1), it is possible to recommend a product that the user is interested in. In addition, recommendation based on attributes and attribute values can provide information about products that are more interesting and more likely to be probable, so that effective recommendations can be made. [0136] (2) For example, in the embodiment, the recording unit 103 determines the category to be processed based on the number of products belonging to each category. However, the specifying unit 102 may determine the classification based on the number of products belonging to each category. The category of its own processing object. That is, the specific unit 102 may refer to the category with only a certain number of products from the search query data. [0137] (3) For example, the category may not be a tree structure. That is, categories may not be hierarchical, and all categories are treated in the same row. At this time, as in the embodiment, it is only necessary to determine the relationship between the users based on the selection result of the search result of the keyword search by the user. [0138] For example, in the embodiment, a case where a combination of an attribute and an attribute value is used as the search information is explained. However, only the attribute may be extracted and used as the search information. In addition, the attributes are not particularly relevant, and the products in the combination of related categories are extracted as common keywords and used as search information. In addition, the combination of categories can also be used as search information. At this time, for example, if the user selects the so-called "wine" category, only related products of the category "red wine" and "white wine" may be listed in the search result. [0139] For example, the case where the combination of attributes and attribute values is extracted from the description in the form of a table has been described, but the combination of attributes and attribute values may be extracted only from the description that is freely entered in the text. Similarly, the combination of attributes and attribute values may not be extracted from the semi-structured description of the text. In addition, for example, it is not necessary to deal with the attributes of the general contract, and different character strings may be formed into different attributes. In addition, for example, a category that does not have a fixed number of products may be treated as a processing target, but a whole category or a category designated by a manager may be used as a processing target. [0140] For example, the acquisition unit 101, the specific unit 102, the recording unit 103, and the display control unit 104 may be implemented by the user terminal 20, respectively. At this time, these functions are mainly realized by the control unit 11. The acquisition unit 101 may be obtained by the server 10 through the network, and the specific unit 102, the recording unit 103, and the display control unit 104 may perform processing based on the data.

[0141]

1‧‧‧ Commodity Information Display System

10‧‧‧Server

11‧‧‧Control Department

12‧‧‧Memory Department

13‧‧‧ Ministry of Communications

20‧‧‧user terminal

21‧‧‧Control Department

22‧‧‧Memory Department

23‧‧‧ Ministry of Communications

24‧‧‧Operation Department

25‧‧‧Display

100‧‧‧Data Memory Department

101‧‧‧Acquisition Department

102‧‧‧Special Department

103‧‧‧Recording Department

104‧‧‧Display Control Department

G1‧‧‧Home

G2‧‧‧Search result screen

G3‧‧‧Product Page

G4‧‧‧Product Page

G5‧‧‧Attribute display screen

G6‧‧‧Search result screen

G7‧‧‧Search result screen

[0021] FIG. 1 is a diagram showing the overall configuration of a product information display system. FIG. 2 is a diagram showing how a user searches for a product. FIG. 3 is a diagram showing how attributes and attribute values are provided to a user. FIG. 4 is a functional block diagram showing an example of functions realized by a product information display system. Figure 5 is a diagram showing an example of category definition data. FIG. 6 is a diagram showing an example of product information. FIG. 7 is a diagram showing an example of retrieval data. Figure 8 shows an example of PV seed data. FIG. 9 is a diagram showing an example of metadata of an expanded product. FIG. 10 is a flowchart showing metadata generation processing. FIG. 11 is a flowchart of display information display processing.

Claims (11)

A product information display system is characterized by including: acquisition means for acquiring product data indicating each category of a plurality of products that can be searched by a keyword; and specific means for keywords input by a plurality of users, The category of the product selected from the search results is specified based on the aforementioned product information; the recording means is that if the same keyword is specified in the plural category by the aforementioned specific method, the plural products included in the plural category Each of the product descriptions is extracted and recorded in the memory means by the attributes and attribute values thereof; and a display control means for displaying information on the display means based on the attributes and the attribute values. For example, the product information display system of the scope of patent application, wherein the aforementioned product description includes at least one of a description in the form of a table and a semi-structured description in the form of a text, and the aforementioned recording means is the aforementioned table in the aforementioned product description At least one of the description of the form and the description of the semi-structured form described above extracts the aforementioned attributes and the aforementioned attribute values. For example, the product information display system of the second patent application scope, The foregoing product description additionally includes: a description freely entered in the form of the text, and the aforementioned recording means is based on the foregoing attributes and the foregoing attribute values extracted from at least one product, and the foregoing attributes and the foregoing are extracted from the foregoing freely entered descriptions of other articles in the form of the text Property value. For example, the product information display system according to any one of claims 1 to 3, in which the aforementioned recording means judges whether different plural plural aforementioned attributes have the same meaning as each other, and judges the aforementioned plural attribute having the same meaning. Unify. For example, the product information display system of the fourth scope of the patent application, wherein the aforementioned recording means determines the plural attribute by determining whether or not the similarity between the plural attributes according to the number of types of common attribute values is greater than a threshold value. Whether they have the same meaning as each other. For example, the product information display system according to any one of claims 1 to 3, wherein the display control means provides a screen for each user to select at least one of the aforementioned attribute and the aforementioned attribute value, so that The aforementioned information of the selection result on this screen is displayed on the aforementioned display means. For example, the product information display system according to any one of claims 1 to 3, wherein the display control means searches for at least one of the attribute and the attribute value of the product requested by each user to display the product page. Matching other products, including links to other searched products On the product page, the information is displayed on the display means. For example, the product information display system according to any one of claims 1 to 3, in which the aforementioned specific means or the aforementioned recording means determine the category to be processed according to the number of products belonging to each category. For example, the product information display system according to any one of the scope of claims 1 to 3, in which each product belongs to a category defined by a tree structure, and the aforementioned recording means is to use the same specific keywords for the same keywords When a category of a plurality of sibling nodes is specified, a combination of categories of the plurality of sibling nodes is recorded. When the category of the parent node of the combination is selected, the display control means displays the information on the display means. A product information display method is characterized in that a computer executes the following steps: an acquisition step that acquires product data representing each category of a plurality of products that can be searched by a keyword; and a specific step that is entered separately by a plurality of users Keywords, according to the aforementioned product data to specify the category of the product selected from the search results; the recording step, if the same keyword is specified in the plural category by the aforementioned specific step, the classification based on the plural category Each product description of the plurality of products included is extracted and recorded in the record A memory means; and a display control step for displaying information on the display means according to the aforementioned attribute and the aforementioned attribute value. A program product that enables a computer to function as a means: acquisition means that acquires product data indicating each category of a plurality of products that can be searched by a keyword; specific means that distinguishes a plurality of users from each other The input keyword specifies the category of the product selected from the search results based on the aforementioned product information; the recording means is that if the same keyword is identified by plural categories for the same keyword by the aforementioned specific means, according to the plural category The product description of each of the plurality of included products is extracted by the attributes and attribute values of the product and recorded in the memory means; the display control means is configured to display information based on the attributes and the attribute values on the display means.