Server management and product search methods

Abstract

This invention is a technology that recommends products that match a user's preferences, as well as similar or related products, based on the user's continuous usage. This invention achieves the above function through the following relationship: A management server 120 is connected to the user terminal 100 and the server 110 via a network. It includes a relevance coefficient information parsing unit 150 that analyzes the relevance coefficients of products based on search patterns contained in the search history 210 provided by the user information storage unit 130 and extracts related product information of the analyzed products from the server 110; and a search unit 160 that extracts product information with a relevance level of 1 corresponding to the acquired user information 200 from the personal product database 140.

Description

Technical Field

[0001] This invention relates to a management server and a product search method. Background Technology

[0002] With the advancement of information technology, people can more easily search for and purchase products that match their preferences from countless goods around the world. However, people's preferences vary. For example, those who are not knowledgeable about interior design may find it difficult to find products that suit their tastes. Or even those who are very knowledgeable about interior design may find it difficult to find products that match their preferences from those they are unfamiliar with.

[0003] Prior art literature

[0004] Patent documents

[0005] Patent Document 1: Japanese Patent Application Publication No. 2007-269491 Summary of the Invention

[0006] The problem that the invention aims to solve

[0007] For example, Patent Document 1 discloses a push mechanism that quantifies and records personal preferences and characteristics of red wine as a commodity, and automatically selects and matches red wines that match personal preferences.

[0008] However, in Patent Document 1, once personal preference information is registered in the system, users can only change the relevant information manually. In other words, because previous recommendation systems lacked learning capabilities, even with continued use, they could not be developed into systems capable of recommending products that match user preferences.

[0009] The purpose of this invention is to provide a technology that, through continuous use, can recommend products that match user preferences, as well as similar or related products.

[0010] Methods for solving problems

[0011] In the invention presented in this case, a representative summary will be briefly described, the specific details of which are as follows.

[0012] This invention is implemented via a user terminal, a server storing large amounts of data, and a management server connected to a network. The invention includes an information analysis unit capable of analyzing product relevance coefficients based on search patterns containing search history information provided by a user information storage unit, and extracting relevance coefficients from a server that stores large amounts of related product information of the analyzed products. This relevance coefficient information analysis unit updates the analyzed product relevance coefficients by overwriting them with the relevance coefficient information contained in the extracted product information. The management server has a personal product database storing the updated product information. Furthermore, the management server has a search unit that extracts product information with a relevance level of 1 corresponding to the acquired user information from the personal product database. This search unit generates a search interface based on attribute data and images from the extracted product information. Additionally, this search unit extracts product information with a similarity level of 1 from the personal product database that corresponds to the target benchmark product determined by the user dragging and dropping desired products in the search interface displayed on the user terminal, and the additional conditions determined by the user inputting filter tags received from the user in the search interface. The search unit generates the search interface based on the extracted product information. Furthermore, this search unit can provide the generated search interface to the aforementioned user terminal.

[0013] Invention Effects

[0014] This invention can push products that match, are similar to or related to the user's preferences through continuous use. Attached Figure Description

[0015] Figure 1 This is a schematic diagram illustrating an example configuration of a product search system including a management server in one embodiment of the present invention.

[0016] Figure 2 This is a schematic diagram illustrating the user information stored in the user information storage unit of the management server in one embodiment of the present invention.

[0017] Figure 3 This is a schematic diagram illustrating the overall processing in one embodiment of the present invention.

[0018] Figure 4 This is a schematic diagram illustrating user information acquisition processing and user information storage processing in one embodiment of the present invention.

[0019] Figure 5 This is a schematic diagram illustrating product information and correlation coefficient information in one embodiment of the present invention.

[0020] Figure 6 This is a schematic diagram illustrating the correlation coefficient information parsing and processing in one embodiment of the present invention.

[0021] Figure 7 This is a schematic diagram illustrating similarity and relevance in one embodiment of the present invention.

[0022] Figure 8 This is a schematic diagram illustrating an example of the configuration of a search interface displayed on a user terminal according to one embodiment of the present invention.

[0023] Figure 9 This is a schematic diagram illustrating an example of the configuration of a search interface displayed on a user terminal according to one embodiment of the present invention.

[0024] Figure 10 This is a schematic diagram illustrating the target search process in one embodiment of the present invention.

[0025] Figure 11 This is a schematic diagram illustrating the re-search process in one embodiment of the present invention.

[0026] Figure 12 This is a schematic diagram illustrating an example of the configuration of a search interface displayed on a user terminal according to one embodiment of the present invention.

[0027] Figure 13 This is a schematic diagram illustrating a similar search process in one embodiment of the present invention.

[0028] Figure 14 This is a schematic diagram illustrating an example of the configuration of a search interface displayed on a user terminal according to one embodiment of the present invention.

[0029] Figure 15 This is a schematic diagram illustrating a search process in one embodiment of the present invention.

[0030] Figure 16 This is a schematic diagram illustrating an example of the configuration of a search interface displayed on a user terminal according to one embodiment of the present invention.

[0031] Figure 17 This is a schematic diagram illustrating a single-item search process in one embodiment of the present invention.

[0032] Figure 18 This is a schematic diagram illustrating an example of the configuration of a search interface displayed on a user terminal according to one embodiment of the present invention.

[0033] Figure 19 This is a schematic diagram illustrating the purchase interface processing in one embodiment of the present invention. Detailed Implementation

[0034] The embodiments of the present invention will be described below with reference to the accompanying drawings. This description is merely an example and is intended to explain the invention only, not to limit it. Furthermore, in the entire set of drawings illustrating the embodiments, the same reference numerals will generally be used for the same parts, and repeated descriptions will be omitted.

[0035] System Composition

[0036] Figure 1 This is a schematic diagram illustrating an example of the configuration of a product search system according to one embodiment of the present invention.

[0037] like Figure 1 As shown, the product search system consists of a user terminal 100, a management server 120 connected to the user terminal 100 via a network, and a server 110 for storing large amounts of data connected to the management server 120 via a network. In other words, the management server 120 is connected to the user terminal 100, the server 110, and the network.

[0038] User terminal 100 refers to a device terminal owned by the user, such as a smartphone or computer. Server 110, which stores large amounts of data, shares this data via network connection. For example... Figure 1 As shown, the server 110 for storing large amounts of data can be connected to the outside of the management server 120. Alternatively, the server 110 for storing large amounts of data can also be contained within the management server 120.

[0039] The management server 120 includes a user information storage unit 130, a personal product database 140, a correlation coefficient information analysis unit 150, a search unit 160, a purchase interface generation unit 170, and a purchase interface provision unit 180.

[0040] User information storage unit 130 stores user information (described later). Figure 2 The personal product database 140 stores product information (described later). Figure 5 ).

[0041] The correlation coefficient information analysis unit 150 analyzes product correlation coefficients based on the search patterns contained in the search history 210 provided by the user information storage unit 130. The correlation coefficient information analysis unit 150 first extracts the product information for each product to be analyzed from the big data storage server. Then, it updates the product correlation coefficients by overwriting them with the correlation coefficient information contained in the extracted product information. Finally, the correlation coefficient information analysis unit 150 sends the product information containing the updated correlation coefficient information to the personal product database 140.

[0042] The search unit 160 generates a search interface 700 (described later). Figure 7 ).

[0043] The purchase interface generation unit 170 generates the purchase interface 1800. First, the purchase interface generation unit 170 extracts product information about the desired product sent by the user terminal 100 from a server with large data storage. Then, the purchase interface generation unit 170 generates the purchase interface 1800 based on attribute data (details described later), images, and specifications from the extracted product information.

[0044] The purchase interface providing unit 180 provides the purchase interface 1800 to the user terminal 100. Then, the user terminal 100 displays the provided purchase interface 1800.

[0045] User Information Storage Department

[0046] Figure 2 (a) is a schematic diagram illustrating a configuration example of user information 200 stored in the user information storage unit 130 of the management server 120 in one embodiment of the present invention.

[0047] like Figure 2 As shown in (a), user information 200 consists of data items such as user ID, name, contact information, friend information, and search history 210.

[0048] The User ID is a symbol used by the management server (120) to identify users. The Name field represents the user's name. Contact information includes the user's email address and phone number, etc. Search history (210) represents the user's search history. (Details follow)

[0049] Friend information refers to the names, addresses, and contact information of recipients to whom a user purchases goods using this invention and then sends them as gifts via the internet (a well-known method omitted here). By including friend information in user information 200, the user's gift-giving history and friend characteristics (age, gender, etc.) can be digitized and reflected in search criteria.

[0050] Figure 2 (b) is a schematic diagram illustrating an example of the configuration of a search history record 210 contained in user information 200 stored in the user information storage unit 130 of the management server 120 according to one embodiment of the present invention. The search history record 210 is used to display information such as user ID, date and time, search mode, search criteria, and purchased goods.

[0051] "Date and Time" refers to the date and time when the management server 120 received the user input information. "Search Pattern" refers to the number of times each search category corresponds to a search, as contained in the user information 200 of the user information received by the management server 120 from the user input information.

[0052] "Search criteria" are the product information of the search objects corresponding to each search, contained in the search history records 210 of the user information 200 that the management server 120 receives from the user input information. For example, search criteria are composed of "type", "category", "product name", "brand", "features" and "price".

[0053] Comprehensive handling

[0054] Figure 3 This is a schematic diagram illustrating the overall process in one embodiment of the present invention.

[0055] First, in S301, the management server 120 will perform user information retrieval processing (in Figure 4 (a) (described later). Then, in S302, the management server 120 will perform user information storage processing (in... Figure 4 (a) (described later). In addition, the S302 process will only run when the user uses the system of the present invention for the first time.

[0056] In S303, the management server 120 will perform correlation coefficient information parsing and processing (in Figure 6 (To be continued later).

[0057] In S304, the management server 120 will perform a search process (in Figures 8 to 17 (To be described later). Then, in S305, the management server 120 will ask the user to determine whether to search again. If the user determines to search again, in S306, the management server 120 will perform the search again (in... Figure 11 (To be continued later).

[0058] If the user decides not to perform another search, in step S307, the management server 120 will ask the user to decide whether to purchase the item. When the user decides to purchase the item, in step S308, the management server 120 will provide the purchase interface. Figure 19 (To be continued later). If the user decides not to purchase the product, the management server 120 will terminate this program.

[0059] User information acquisition and storage processing

[0060] Figure 4 (a) is a schematic diagram illustrating user information acquisition processing and user information storage processing in one embodiment of the present invention. The method by which the management server 120 acquires user information 200 will be described below.

[0061] First, in S401, the management server 120 provides the user information input interface to the user terminal 200. Second, in S402, the user terminal 200 displays and plays the user information input interface.

[0062] Secondly, when a user uses the system of this invention for the first time, in S403, the user terminal 100 receives the input content of user information 200 from the user. Furthermore, user information 200 consists of data items such as user ID, name, contact information, and search history records 210. Then, in S404, the user terminal 100 sends the user information 200 received in S403 to the relationship server 120.

[0063] Next, in S405, the management server 120 obtains the user information 200 by receiving the user information 200 sent in S404. Then, in S406, the management server 120 adds a user ID to the received user information 200, and in S403, it stores the user information 200 containing the input content in the user information storage unit 130.

[0064] User Information Acquisition and Processing

[0065] Figure 4 (b) A schematic diagram illustrating the user information acquisition process in one embodiment of the present invention. Hereinafter, a method for the management server 120 to acquire user information 200 will be described.

[0066] When a user uses the system of this invention for the second time or subsequent times, in S407, the user terminal 100 receives the user ID entered by the user. The user ID is also contained in the user information 200. Then, in S408, the user terminal 100 sends the user ID received in S407 to the management server 120.

[0067] Secondly, in S409, the management server 120 retrieves the user information 200 corresponding to the user ID and key sent in S408 from the user information storage unit 130.

[0068] Correlation coefficient information analysis and processing

[0069] Figure 5 This is a schematic diagram illustrating product information and correlation coefficient information in one embodiment of the present invention. Additionally, Figure 6 This is a schematic diagram illustrating the correlation coefficient information parsing and processing in one embodiment of the present invention. The following is for... Figures 5-6 This document describes the method for parsing correlation coefficient information for the management server 120.

[0070] Figure 5(a) shows a summary of the product information. The product information consists of "attribute data", "images", "specifications", and "correlation coefficient information", etc.

[0071] The attribute data consists of "type", "category", "product name", "brand", "features", and "price". Additionally, features are the characteristics that constitute the product, such as "color", "size", and "material".

[0072] For example, the attribute data of product information 502 for product A (501) is as follows: type is "interior decoration", category is "pillow", product name is "product A", color is "gray", etc. Similarly, product D (505) has inherent product information 506.

[0073] "Image" refers to image data representing the appearance of a product, such as JPG data. "Specifications" refers to text data representing the product's specifications.

[0074] "Relevance coefficient information" includes information such as user search patterns and related links. Furthermore, the correlation coefficient is a value ranging from 0 to 1, representing the degree of relevance between products. The correlation coefficient is determined by the relative degree of association between each product and other products. A correlation coefficient of 0 indicates the lowest relevance, while 1 indicates the highest relevance. The correlation coefficient increases from 0 to 1. Additionally, even for the same product, the correlation coefficient can vary for different users based on factors such as age, gender, and region. Figure 5 (b) is a summary diagram of the correlation coefficient information.

[0075] Figure 5 (b) A search pattern 510 is shown when a user searches for products, starting with product A (501) and proceeding to product B, product C, and product D (505). For example, the search pattern 510 at this time includes a route 511 from product A to product B.

[0076] in addition, Figure 5(b) shows the state reflecting the search pattern 510 and the generated correlation coefficient links 520. For each product (node), a correlation coefficient is defined between the routes connecting each product (node). Furthermore, based on the search pattern 510, the correlation coefficient information parsing unit 150 can increase the correlation coefficients of each product (node). More specifically, the correlation coefficient information parsing unit 150 can increase the correlation coefficients of the routes (links connecting products) included in the search pattern 510. According to the above, the correlation coefficients of the searched routes (links connecting products) can increase. For example, if product C and product D have different colors and sizes, the correlation coefficient between product C and product D before the search is "0.5". Additionally, the correlation coefficient 521 between any products not included in route 511 is "0.4". At this time, the correlation coefficient 522 between product C and product D included in route 511 connecting product C and product D after the search increases from "0.5" to "0.7". In other words, the relevance coefficient information analysis unit 150 adds a weight vector to the relevance of the product based on the searched content, causing the relevance coefficient 522 to increase. However, the relevance coefficient 521, which is not included in path 511, cannot have its weight vector added, and its relevance coefficient remains unchanged at "0.4".

[0077] like Figure 6 As shown, firstly, in S601, the correlation coefficient information parsing unit 150 extracts the search history 210 corresponding to the user ID to be parsed from the user information storage unit 130. The search history 210 consists of information such as user ID, date and time, search pattern, search conditions, and purchased products. Then, the correlation coefficient information parsing unit 150 provides the search history 210 extracted from the user information storage unit 130 to the correlation coefficient information parsing unit 150. Then, in S603, the correlation coefficient information parsing unit 150 parses the correlation coefficient of the products based on the search pattern contained in the search history 210 provided above. Specifically, the correlation coefficient information parsing unit 150 connects to each product constituting the route 511 contained in the search pattern 510 and parses the correlation coefficient by extracting the (specific) correlation coefficient information inherent to each user.

[0078] Next, in S604, the correlation coefficient information parsing unit 150 extracts the product information of each product to be parsed from the server 110 storing big data. Specifically, the correlation coefficient information parsing unit 150 is connected to each product and parses the correlation coefficient by extracting the common (general) correlation coefficient information shared by all users. Then, in S605, the correlation coefficient information parsing unit 150 updates itself by overwriting the correlation coefficient information contained in the extracted product information with the correlation coefficient of the parsed product. In other words, the correlation coefficient information parsing unit 150 reflects the search pattern 510 obtained through the search performed by the user and updates itself by overwriting the general correlation coefficient information extracted from the server 110 storing big data with the specific correlation coefficient information inherent to each user.

[0079] Then, in S606, the product information containing the updated correlation coefficient information is sent to the personal product database 140. Next, in S607, the personal product database 140 stores the updated product information. Furthermore, the product information consists of attribute data and correlation coefficient information, etc. The personal product database 140 updates and stores the correlation coefficient information in the product information.

[0080] Similarity and relevance

[0081] Figure 7 This is a schematic diagram illustrating similarity and relevance in one embodiment of the present invention.

[0082] like Figure 7 As shown in (a), the products are classified according to the similarity of any product baseline. Similarity is determined by the number of items in the product information attribute data that have the same or similar parameters, or by the number of items in the product information attribute data that have the same or similar parameters. In other words, similarity represents the number of items in the product's attribute data (attribute representation) that have the same or similar parameters. For example, when the similarity is 1, the number of items with the same or similar parameters is one. If similarity n+1 is always included in similarity n, then the higher the similarity value, the higher the similarity between products.

[0083] For example, using product A (700) as the base, products are classified according to similarity 1 (710), similarity 2 (720), and similarity 3 (730). Similarity 1 includes products of the same type as product A at the base. When product A is of type "interior decoration," product 711 is included in similarity 1 because it is also an "interior decoration" type product. In other words, products with one item having the same or similar parameters as the base product have a similarity of 1.

[0084] Similarity level 2 includes goods that belong to the same category as the base item A. When the category of item A is "cushion," item 721 is included in similarity level 2 because it is also a "cushion" item. In other words, items with two items having the same or similar parameters as the base item have a similarity level of 2. Similarity level 3 includes items of the same color as the base item A. When the color of item A is "gray," item 731 is included in similarity level 3 because it is also a "gray" item. In other words, items with three items having the same or similar parameters as the base item have a similarity level of 3.

[0085] As described above, by categorizing products based on their similarity, the search unit 160 can periodically filter out highly similar products for the user. In other words, the search unit can provide targeted searches (for...) Figure 10 (and so on, as will be discussed later) and similar searches.

[0086] In the target search, the search unit 160 extracts product information with a "similarity level of 1" from the personal product database 140 that corresponds to the user's desired target benchmark product 810 and additional conditions (details described later). The resulting search interface 840 is then provided to the user.

[0087] In similar searches, the search unit 160 extracts product information with a "similarity level 2" from the personal product database 140 that corresponds to multiple target benchmark products 1210 and additional conditions (details below) as desired by the user. The resulting search interface 1240 is then provided to the customer.

[0088] like Figure 7 As shown in (b), each product is categorized based on its relevance to any given product. Relevance is determined by the correlation coefficient contained in the product information. A higher correlation coefficient indicates a stronger relevance between products. Furthermore, the more searches conducted, the higher the correlation coefficient becomes, and the stronger the relevance increases.

[0089] For example, using product A (700) as the base, products are classified according to relevance 1 (750), relevance 2 (760), and relevance 3 (770). As another example, relevance 1 includes products with a relevance coefficient to the base product A of 0.1 or higher and lower than 0.3. Product 751 has a relevance coefficient of 0.2 with product A, and is therefore included in relevance 1.

[0090] Relevance level 2 includes goods with a correlation coefficient of 0.3 or higher and less than 0.7 with the base point good A. For example, good 761 has a correlation coefficient of 0.4 with good A, so it is included in relevance level 2. Relevance level 3 includes goods with a correlation coefficient of 0.7 or higher and less than 1.0 with the base point good A. For example, good 761 has a correlation coefficient of 0.8 with good A, so it is included in relevance level 3.

[0091] By classifying products according to their relevance as described above, the search unit 160 can refer to this relevance and filter out highly relevant products for users at different stages, that is, provide combination search (in... Figure 15 (The following will be discussed) service.

[0092] In the matching search, the search unit 160 extracts highly relevant product information from the personal product database 140 that corresponds to the user's desired matching benchmark product 1510 and additional conditions (details described later). The resulting search interface 1440 is then provided to the customer.

[0093] Target search

[0094] Search interface

[0095] Figure 8 , Figure 9 This is a schematic diagram illustrating an example of the configuration of a search interface displayed on a user terminal 100 according to one embodiment of the present invention.

[0096] like Figure 8 As shown in (a), the search interface 800 includes an icon 810 displaying the "target reference product". The target reference product 810 is the product used as a base point when the search unit 160 performs a search. The target reference product 810 is displayed in the center cell of a 3×3 grid. Therefore, icons 831 to 838 of "recommended products" are displayed at eight locations: above, below, left, right, upper left, upper right, lower left, and lower right corners of the target reference product 810. Eight icons 831 to 838 of "recommended products" are displayed, surrounding the target reference image 810. In addition, the target reference product 810 is determined by accepting input content from the user dragging from "any cell other than the center cell" to the center cell.

[0097] Additionally, when the search interface 800 receives input from any cell outside the center cell dragged to the center cell, it replaces the icons displayed in the cells outside the center cell with the icons displayed in the center cell. In other words, the center cell displaying the 3x3 grid, along with the eight "recommended product" icons 831-838 above, below, left, right, upper left, upper right, lower left, and lower right of it, will always display product information; there will be no empty slots.

[0098] The search interface 800 includes a magnified display label 811 for the target reference product 810. The image of the target reference product 810 is magnified and displayed in the magnified display label 811. The search interface 800 generates a search unit 160, which is then displayed on the user terminal 100.

[0099] The search interface 800 includes filter tags 812. Filter tags 812 allow you to set product "category" and "features" as additional search criteria.

[0100] The search interface 800 includes a target search label 813. The target search label 813 can perform a target search by accepting user input. Figure 10 (To be continued).

[0101] The search interface 800 includes a re-search tab 820. The re-search tab 820 can perform a re-search by accepting user input. Figure 11 (To be continued).

[0102] The search interface 800 includes a previous screen display tab 821. The previous screen display tab 821 causes the user terminal 100 to display the previous screen based on the user's actions. Furthermore, "previous screen" refers to, for example, the screen "before switching" when clicking the search tab 820 to switch interfaces.

[0103] The search interface 800 includes keyword search tags 822. Keyword search tags 822 can directly search for the products the user wants by accepting the input of specific keywords.

[0104] The search interface 800 contains icons for multiple "recommended products" from 831 to 838. These recommended products are generated by the search unit 160 by extracting product information from the personal product database that corresponds to the user's age, gender, etc., based on the user information obtained through user information processing (S301 or S302). Furthermore, icons 831 to 838 are randomly displayed in eight cells of a 3x3 table, excluding the center cell.

[0105] like Figure 8 As shown in (b), the search interface 840 includes target benchmark product 810, re-search label 820, previous interface display label 821, and keyword search label 822, etc.

[0106] The search interface 840 contains icons for "recommended products" from 841 to 848. Recommended products are, in other words, "search result products." That is, icons 841 to 848 represent products from the search results obtained by the search unit 160 after performing a target search. Furthermore, icons 841 to 848 are randomly displayed in eight cells of a 3x3 table, excluding the center cell.

[0107] In addition, the search interface 840 is displayed on the user terminal 100 after the user clicks the search tag 820 (by initiating a similar search process) in the search interface 800.

[0108] In addition, when the management server 120 receives input from the user terminal 100 via the search tag 820, it can activate the charging function through banner ads for the products corresponding to the nine icons displayed in the search interface 800.

[0109] As described above, since the re-search 820 has the function of banner advertising, the management server 120 can provide the benefits of using the present invention to the sellers of goods.

[0110] like Figure 9 As shown in (a), the search interface 850 includes a target benchmark product 810, a re-search tag 820, and a previous interface display tag 821, etc.

[0111] The search interface 850 contains icons for "recommended products" from 851 to 858. Recommended products are, in other words, "search result products." That is, icons 851 to 858 represent products from the search results obtained by the search unit 160 during a target search. Furthermore, icons 851 to 858 are randomly displayed in eight cells of a 3x3 table, excluding the center cell.

[0112] Search screen 850 includes a "Pin Products" option icon 900. Pin products refer to the products the user selects from the pushed products 852 to 959 that they want to be displayed consistently (e.g., products 851 and 856). In other words, as... Figure 9 As shown in (b), items 851 and 856 attached to the fixed item option icon 900 will remain fixed even after the user presses the search bar 820. In other words, the items selected as fixed items and displayed in the fixed item option icon 900 will not be replaced even after the search bar 820 is pressed. Furthermore, icon 900 is determined by the user's long press on the icon they wish to fix as a fixed item for a certain period of time (e.g., three seconds).

[0113] like Figure 9 As shown in (b), the search interface 860 includes a target benchmark product 810, a re-search tag 820, and a previous interface display tag 821, etc.

[0114] The search interface 860 contains icons for "recommended products" from 861 to 868. Recommended products are, in other words, "search result products." That is, icons 861 to 868 represent products from the search results obtained by the search unit 160 during a target search. Furthermore, icons 861 to 868 are randomly displayed in eight cells of a 3x3 table, excluding the center cell.

[0115] Additionally, the search interface 860 includes a "Fixed Product" option icon 900. This means that products 861 and 866 with the "Fixed Product" option icon 900 will be displayed as products 851, 856, and each other identically listed product until the user clicks the search tag 820.

[0116] Target search processing

[0117] Figure 10 This is a schematic diagram illustrating a target search process according to one embodiment of the present invention. Hereinafter, the method by which the search unit 160 generates search interfaces 800 and 840 will be described.

[0118] First, in S1001, the search unit 160 extracts product information with a "relevance score of 1" corresponding to age, gender, etc., from the personal product database 140, based on the user information 200 obtained in S301 or S302. Furthermore, a product with a relevance score of 1 refers to, for example, a product whose correlation coefficient with the base product A is equal to or higher than 0.1 and lower than 0.3. Next, in S1002, the search unit 160 generates a search interface 800 based on the extracted product information.

[0119] Specifically, the search unit 160 generates a search interface 800 based on the attribute data, including "product name," "brand," "features," "price," and images, contained in the extracted product information.

[0120] Next, in S1003, the search unit 160 provides the generated search interface 800 to the user terminal 100. Then, in S1004, the user terminal 100 displays the search interface 800 provided above on the player.

[0121] In S1005, the user terminal 100 determines the target benchmark product 810 by receiving input from the user dragging and dropping desired products in the search interface 800. Alternatively, in S1006, the user terminal 100 determines additional conditions after receiving input from the user on the filter tags 812 in the search interface 800. Additional conditions refer to filtering conditions where the user adds and sets "type" or "features" to narrow down the search scope when searching for products. Then, in S1007, the user terminal 100 sends the determined target benchmark product 810 and additional conditions to the search unit 160.

[0122] Then, in S1008, the search unit 160 extracts product information with a "similarity of 1" corresponding to the received target benchmark product 810 and additional conditions from the personal product database 140. Furthermore, the product information is composed of attribute data and correlation coefficient information, etc. The search unit 160 extracts product information corresponding to the aforementioned products from the personal product database 140. Then, in S1009, the search unit 160 generates a search interface 840 based on the extracted product information.

[0123] Specifically, the search unit 160 generates a search interface 840 containing the above content based on the attribute data such as "product name", "brand", "features", "price" and images contained in the extracted product information.

[0124] Next, in S1010, the search unit 160 provides the generated search interface 840 to the user terminal 100. Then, in S1011, the user terminal 100 displays the search interface 840 provided above on the player.

[0125] As described above, by running a target search, the management server 120 can make recommendations based on certain specific product characteristics.

[0126] Re-search processing

[0127] Figure 11 (a) is a schematic diagram illustrating the re-search process in one embodiment of the present invention. The method by which the search unit 160 generates the search interface 850 will be described below.

[0128] First, in S1101, the user terminal 100 asks the user to determine whether to change the target base product 810 or additional conditions. When an instruction to change the target base product 810 or additional conditions is received, the user terminal 100 and the management server 120 repeat the process in S1005 to S1011. When an instruction not to change the target base product 810 or additional conditions is received, the user terminal 100 proceeds to the process in S1102.

[0129] In S1102, the user terminal 100 receives an instruction from the user to press the re-search tag 820, and begins a re-search while keeping the target benchmark product 810 or additional conditions unchanged. Additional conditions refer to filtering conditions that allow the user to add and set "type" or "features" to narrow down the search scope when searching for products. Next, in S1103, the search unit 160 extracts products from the personal product database 140 that maintain a "similarity 2" corresponding to the target benchmark product 810 and the additional conditions, and removes product information already provided to the user terminal 100. For example, if the additional condition is "color," the search unit 160 extracts products with the same "color" from the personal product database 140 and removes product information already provided to the user terminal 100. Furthermore, similarity 2 includes products in the same category as the maintained target benchmark product 810. For example, if the maintained target benchmark product 810 is categorized as "cushion," products that are also "cushion" are included in similarity 2. In other words, if two items have the same or similar parameters as the target benchmark product 810, their similarity is 2. Then, in S1104, the search unit 160 generates a search interface 850 based on the extracted product information.

[0130] Specifically, the search unit 160 generates a search interface 850 containing the above information based on the attribute data included in the extracted product information, such as "product name", "brand", "features", "price" and images.

[0131] Then, in S1105, the search unit 160 provides the generated search interface 850 to the user terminal 100. Then, in S1106, the user terminal 100 displays the search interface 850 provided above on the player.

[0132] Then, in S1107, the user terminal 100 will ask the user to determine whether to perform another search. When an input command to perform another search is received, the user terminal 100 and the management server 120 will repeat the procedures of S1101 to S1107. If no further search is performed, the user terminal 100 will terminate this procedure.

[0133] Furthermore, in the nth re-search S1103, the "similarity n+1" process is handled. That is, in S1107, after the user terminal 100 asks the user to decide whether to perform a re-search, upon receiving a re-search instruction, the procedures of S1101 to S1102 are repeated first. Additionally, the system automatically identifies this re-search as a "second re-search." In other words, the similarity increases with each re-search. Next, the search unit 160 repeats the procedure of S1103. At this time, it extracts products from the personal product database 140 that maintain a "similarity of 3" corresponding to the target reference product 810 and the additional conditions, excluding the product information already provided to the user terminal 100. The reason for selecting "similarity 3" is that the re-search is a "second re-search." Furthermore, similarity 3 includes products that are the same color as the maintained target reference product 810. For example, if the maintained target reference product 810 is "gray," products that are also "gray" are included in similarity 3. In other words, a similarity level of 3 is defined as 3 for products with 3 or more items that are identical or similar to the target benchmark product 810. Then, the search unit 160 repeats the procedures S1104 to S1105. That is, it generates a search interface 850 based on the extracted product information and provides it to the user terminal 100. Afterward, the user terminal 100 repeats the procedures S1106 to S1107.

[0134] Figure 11 (b) is a summary diagram showing the re-search process when a fixed product is selected. The method by which the search unit 160 generates the search interface 860 will be described below.

[0135] In S1108, the user terminal 100 receives the above input by allowing the user to select the fixed product option icon 900 to determine the fixed product. Furthermore, the fixed product is not limited to a single item. Next, in S1109, the user terminal 100 sends the determined fixed product data to the search unit 160. The fixed product data shows the product information that the user wants to display fixedly among the recommended products 850 to 858 (e.g., products 851 and 856). That is, it displays the relevant information of products 851 and 856 with the fixed product option icon 900.

[0136] Then, in S1110, the search unit 160, based on the information contained in the received fixed product data regarding which product among the recommended products 851 to 858 the user wants to display permanently, performs procedures S1103 to S1105 for "cells without a selected fixed product" (that is, cells without a fixed product option icon 900). In this way, the search unit 160 generates a search interface 860 and provides it to the user terminal 100.

[0137] Then, in S1111, the user terminal 100 will display the search interface 860 provided above on the player.

[0138] As described above, by performing a re-search, the management server 120 can recommend products that match the user's preferences or similar or related products by allowing the user to search continuously.

[0139] Similar Searches

[0140] Search interface

[0141] Figure 12 This is a schematic diagram illustrating an example of the configuration of a search interface displayed in a user terminal 100 according to one embodiment of the present invention.

[0142] like Figure 12 As shown in (a), the search interface 1200 includes icons 1210 and 1212 representing "target reference products". Target reference products 1210 and 1212 are used to represent products that serve as base points when the search unit 160 performs similar searches. In addition, either target reference product 1210 or 1212 (e.g., 1210) is displayed in the center cell of a 3×3 table.

[0143] The target reference product 1210 or 1212 (e.g., 1210) is determined by receiving an operation instruction from the user to drag the product from "any cell outside the center cell" to the center cell. The other target reference product 1210 or 1212, besides the one mentioned above, is determined by clicking the desired product after the above drag-to-center-cell operation.

[0144] The search interface 1200 includes a magnified display label 1211 for the target benchmark product 1210. The magnified display label 1211 enlarges the image of the target benchmark product 1210.

[0145] The search interface 1200 includes filter tags 812, target search tags 813, further search tags 820, and the previous screen displays tags 821 and keyword search tags 822.

[0146] The search interface 1200 contains icons for multiple "recommended products" from 1231 to 1238. These recommended products are generated by the search unit 160 by extracting product information with a "relevance score of 1" corresponding to the user's age and gender from the user information 200 obtained through user information retrieval processing (S301 or S302) in the personal product database. Icons 1231 to 1238 are randomly displayed in eight cells of a 3×3 table, excluding the center cell. For example, product 1238 is, as mentioned above, the target reference product 1212, generated after receiving an instruction to drag the target reference product 1210 towards the center cell.

[0147] like Figure 12 As shown in (b), the search interface 1240 includes target benchmark products 1210 and 1212, as well as re-search tags 820, etc.

[0148] Search interface 1240 contains icons for multiple "recommended products" from 1241 to 1247. Recommended products are, in other words, "search result products." That is, icons 1241 to 1247 represent products from search results obtained by the search unit 160 through similar searches. Furthermore, icons 1241 to 1247 are randomly displayed in the cells of a 3×3 table, excluding the selected target benchmark product.

[0149] Secondly, the search interface 1240 is set up in the search interface 1200 by having the user click on the search tag 820 (by initiating a similar search process), which causes the player on the user terminal 100 to switch from the search interface 1200.

[0150] Furthermore, when the management server 120 receives an instruction from the user terminal 100 to click the re-search tag 820, it can also enable the products corresponding to the nine icons displayed on the search interface 1200 to generate banner advertisements. As described above, since the re-search tag 820 has the function of banner advertisements, the management server 120 can also provide the sellers of goods with the benefits of using this invention.

[0151] Similar search processing

[0152] Figure 13 This is a schematic diagram illustrating a similar search process in one embodiment of the present invention. Hereinafter, the method by which the search unit 160 generates search interfaces 1200 and 1240 will be described.

[0153] First, in S1301, the search unit 160 generates a search screen 1200 by running the programs in S1001 to S1004. Next, in S1302, the search unit 160 provides the generated search screen 1200 to the user terminal 100. Then, in S1303, the user terminal 100 displays the aforementioned search screen 1200.

[0154] In S1304, the user terminal 100 determines the target reference product 1210 based on the instruction received from the user in the search interface 1200 to drag the desired product from any cell other than the center cell to the center cell. After completing the drag-to-center operation, the user terminal 100 determines the target reference product 1212 by clicking on the desired product. Then, in S1305, the user terminal 100 determines additional conditions based on the instruction received from the user in the search interface 1200 to input filter tags 812. Finally, in S1306, the user terminal 100 sends the determined target reference products 1210 and 1212, along with the additional conditions, to the search unit 160.

[0155] Next, in S1307, the search unit 160 extracts product information with a "similarity of 1" corresponding to the received multiple target benchmark products 1210, 1212 and additional conditions from the personal product database 140. Then, in S1308, the search unit 160 generates a search interface 1240 based on the extracted product information.

[0156] Specifically, the search unit 160 generates a search interface 1240 based on the attribute data, including "product name", "brand", "features", "price" and image, contained in the extracted product information.

[0157] Then, in S1309, the search unit 160 provides the generated search interface 1240 to the user terminal 100. Then, in S1310, the user terminal 100 displays and plays the aforementioned search interface 1240.

[0158] As described above, by performing similar searches, the management server 120 can make recommendations based on features shared by two or more products.

[0159] Matching search

[0160] Search interface

[0161] Figure 14 This is a schematic diagram illustrating an example of the configuration of a search interface displayed in a player of a user terminal 100 according to one embodiment of the present invention.

[0162] like Figure 14As shown in (a), the search interface 1400 includes an icon 1410 representing "matching reference product". The matching reference product 1410 displays the product that the search unit 160 uses as the base point when performing a matching search. In addition, the matching reference product 1410 is displayed in the center cell of a 3x3 table.

[0163] The matching of the base product 1410 is determined based on the instruction received from the user to drag from "any cell outside the center cell" to the center cell.

[0164] The search interface 1400 includes a magnified display label 1411 for the target benchmark product 1410. The magnified display label 1411 displays an enlarged image of the target benchmark product 1410.

[0165] The search interface 1400 includes a matching tag 1412. The matching tag 1412 can perform matching searches by receiving user click commands. (To be continued...) Figure 15 )

[0166] The search interface 1400 includes filter tags 812, target search tags 813, further search tags 820, previous screen display tags 821 and keyword search tags 822.

[0167] The search interface 1400 contains icons for multiple "recommended products" from 1431 to 1438. These recommended products are generated by the search unit 160 after extracting product information from the user information 200 obtained through user information retrieval processing (S301 or S302) from the personal product database, identifying products with a "relevance score of 1" corresponding to the user's age, gender, etc. Icons 1431 to 1438 are randomly displayed in eight cells of a 3x3 table, excluding the center cell.

[0168] like Figure 14 As shown in (b), the search interface includes target benchmark product 1410 and re-search tags 820, etc.

[0169] The search interface 1400 contains icons for multiple "recommended products" from 1441 to 1448. Recommended products are, in other words, "search result products." That is, icons 1441 to 1448 represent products from the search results obtained by the search unit 160 through a combination search. Furthermore, icons 1441 to 1448 are randomly displayed in cells outside the center cell selected as the basis for the combination search in a 3x3 table.

[0170] Additionally, the search interface 1440 is displayed on the user terminal 100 after the user clicks the search tag 820 (by initiating the matching search process) within the search interface 1400, causing the player to switch from the search interface 1400 to the search interface 1400.

[0171] Furthermore, when the management server 120 receives input from the user terminal 100 clicking the re-search tag 820, it can also enable banner advertising for the products displayed in the search interface 1400 corresponding to the nine-grid icon. As described above, by enabling banner advertising for the re-search tag 820, the management server 120 can provide the benefits of using this invention to product sellers.

[0172] Combined search processing

[0173] Figure 15 This is a schematic diagram illustrating a search process in one embodiment of the present invention. Hereinafter, the method by which the search unit 160 generates search interfaces 1400 and 1440 will be described.

[0174] First, in S1501, the search unit 160 generates a search interface 1400 by running the programs in S1001 to S1004. Next, in S1502, the search unit 160 provides the generated search interface 1400 to the user terminal 100. Then, in S1503, the user terminal 100 displays the aforementioned search interface 1400.

[0175] In S1504, the user terminal 100 determines the matching reference product 1410 based on the received instruction from the user to drag the desired product from any cell outside the center cell to the center cell in the displayed search interface 1400. Then, in S1505, the user terminal 100 determines additional conditions based on the input of the filter tab 812 received from the user in the search interface 1400. Finally, in S1506, the user terminal 100 sends the determined matching reference product 1410 and the additional conditions to the search unit 160. The matching reference product 1410 indicates the product used as the base point when the search unit 160 performs a matching search.

[0176] Next, in S1507, the search unit 160 extracts the received product information that is highly relevant to the matching benchmark product 1510 and additional conditions from the personal product database 140. Furthermore, the product information is composed of attribute data and correlation coefficient information, etc. The search unit 160 extracts the corresponding product information from the personal product database 140. Then, in S1508, the search unit 160 generates a search interface 1440 based on the extracted product information.

[0177] Specifically, the search unit 160 generates the search interface 1440 based on the attribute data "product name", "brand", "features", "price" and image contained in the extracted product information.

[0178] Then, in S1509, the search unit 160 provides the generated search interface 1440 to the user terminal 100. Then, in S1510, the user terminal 100 displays the search interface 1440.

[0179] As described above, by performing a matching search, the management server 120 can recommend suitable matching products from products that are the same type as a specific product but different in category.

[0180] Single item search

[0181] Search interface

[0182] Figure 16 This is a schematic diagram illustrating an example of the configuration of a search interface displayed in a player of a user terminal 100 according to one embodiment of the present invention.

[0183] like Figure 16 As shown in (a), the search interface 1600 includes an icon 1610 for the target benchmark product or a set of benchmark products. Additionally, the search interface 1600 includes a "set product icon" 1638. The set product icon 1638 represents the icon of a single item in a "set product" consisting of multiple items.

[0184] The search interface 1600 contains icons for multiple "recommended products" from 1631 to 1638. These recommended products are generated by the search unit 160 extracting product information with a "relevance score of 1" corresponding to the user's age and gender from the personal product database, obtained through the user information acquisition process (S301 or S302). Icons 1631 to 1638 are randomly displayed in eight cells outside the center cell of a 3x3 table.

[0185] like Figure 14 As shown in (b), the search interface 1640 includes a single benchmark product 1638 and a re-search tag 820, etc.

[0186] Search interface 1640 contains icons for multiple "recommended products" from 1641 to 1648. Recommended products are, in other words, "search result products." That is, icons 1641 to 1648 represent products from the search results obtained by the search unit 160 during individual item searches. Furthermore, icons 1641 to 1648 are randomly displayed in a 3x3 table after removing the center cell of the selected single-item baseline product. The single-item baseline product 1638 is the product used as the base point when the search unit 160 performs individual item searches. Additionally, the user can select the single-item baseline product 1638 by dragging the set product icon 1638 to the center cell in search interface 1600.

[0187] Furthermore, the search interface 1640 is switched to display on the user terminal 100's player by having the user click the search tag 820 in the search interface 1600 (by initiating a single-item search process).

[0188] Furthermore, when the management server 120 receives a click on the re-search tag 820 from the user terminal 100, it can enable the products corresponding to the nine-grid icons displayed in the search interface 1600 to have a banner advertising fee. As described above, by enabling the re-search tag 820 to have a banner advertising fee, the management server 120 can provide the sellers of the products with the benefits of using this invention.

[0189] Single item search processing

[0190] Figure 17 This is a schematic diagram illustrating a single-item search process according to one embodiment of the present invention. Hereinafter, the method for the search unit 160 to generate search interfaces 1600 and 1640 will be described.

[0191] First, in S1701, the search unit 160 generates a search interface 1600 by running the programs in S1001 to S1004. Next, in S1702, the search unit 160 provides the generated search interface 1600 to the user terminal 100. Then, in S1703, the user terminal 100 displays the aforementioned search interface 1600.

[0192] Then, in S1704, the user terminal 100 determines the single-item base item 1638 by receiving the user's operation of dragging the package item icon 1638 from the displayed search interface 1600 to the center grid. Then, in S1705, the user terminal 100 sends the determined single-item base item 1638 to the search unit 160.

[0193] Then, in S1706, the search unit 160 extracts information on eight items that are highly relevant to the individual reference item 1638 included in the package from the personal product database 140. Furthermore, "package" refers to a product where the advertiser pre-selects multiple (eight or more) items and defines them as a single product. Alternatively, the multiple items included in a package may share a common characteristic in some conceptual sense. Then, in S1707, the search unit 160 generates a search interface 1640 based on the extracted product information.

[0194] Specifically, the search unit 160 generates a search interface 1640 based on the attribute data "product name", "brand", "features", "price" and image contained in the extracted product information.

[0195] Then, in S1708, the search unit 160 provides the generated search interface 1640 to the user terminal 100. Then, in S1709, the user terminal 100 displays the aforementioned search interface 1640.

[0196] As mentioned above, by performing single-item searches, the management server 120 can make recommendations by breaking down the concept map of a world view that combines several items into individual items.

[0197] Inspiration Search

[0198] For example, in target search processing (similar to search, combination search and single item search), when neither the target search tag 813 nor the combination tag 1412 in the search screen 800 are selected, the search tag 820 will become the "inspiration tag".

[0199] At this point, the icon for the target benchmark product 810 will be removed from the center cell of the 3x3 table. That is, only the icons 831 to 838 of the eight "recommended products" from the top, bottom, left, right, upper left, upper right, lower left, and lower right of the target benchmark product 810 will be displayed.

[0200] Then, the target benchmark product 810 is determined by receiving an instruction from the user to drag the icons 831-838 of the eight "recommended products" displayed in "any cell outside the center cell" to the center cell.

[0201] Furthermore, the eight icons 831-838 displaying "recommended products" are, for example, not selected from the personal product database 140 based on search criteria. In other words, the eight icons 831-838 displaying "recommended products" are not randomly selected based on search criteria.

[0202] Purchase interface

[0203] Figure 18 This is a schematic diagram illustrating an example of the configuration of a purchase interface 1800 displayed on a user terminal 100 according to one embodiment of the present invention.

[0204] Figure 18 (a) and Figure 8 (a) shows the same content. The search interface 800 includes the target benchmark product 810, the search tag 820, and the tag displayed on the previous interface 821, etc.

[0205] like Figure 18 As shown in (b), the purchase interface 1800 includes a display area 1810 showing the specifications and style of the product to be purchased, a purchase label 1820, and a cancellation label 1830.

[0206] The purchase interface provides processing.

[0207] Figure 19 This is a schematic diagram illustrating the process of providing a purchase interface in one embodiment of the present invention. Hereinafter, a method for providing a purchase interface 1800 to the management server 120 will be described.

[0208] First, in S1901, the user terminal 100 determines the product to be purchased by having the user click on the magnified display label 811 included in the displayed search interface 800, etc.

[0209] Furthermore, when the management server 120 receives an operation instruction from the user terminal 100 to click the zoom-in display label 811 in S1901, it can enable the label to have a banner advertising charging function. As described above, since the zoom-in display label 811 has the function of banner advertising, the management server 120 can provide the sellers of goods with the benefits of using the present invention.

[0210] Then, in S1902, the user terminal 100 sends the data of the desired purchased goods to the management server 120. In S1903, the management server 120 extracts the received product information related to the desired purchased goods from the server with big data storage. In S1904, the management server 120 generates a purchase interface 1800 based on the extracted product information.

[0211] Specifically, the search unit 160 generates a purchase interface 1800 containing the above information based on the attribute data "product name", "brand", "features", "price" and image contained in the extracted product information.

[0212] Then, in S1905, the management server 120 provides the generated purchase interface 1800 to the user terminal 100. Then, in S1906, the user terminal 100 displays the aforementioned purchase interface 1800.

[0213] Then, in S1907, the user terminal 100 prompts the user to decide whether to purchase the desired item. Upon receiving the instruction to purchase, the user terminal 100 instructs the user to click the purchase tab 1820 included in the purchase interface 1800 to confirm the purchase. Alternatively, if no purchase decision is made, the user terminal 100 will terminate the process.

[0214] Effects of this implementation method

[0215] According to the embodiments of the present invention described above, the management server 120 can generate a search interface 800 based on the user information 200 received from the user terminal 100 and the product information received from the server 110, and display and play it, thereby recommending products that are more in line with the user's preferences and similar or related products.

[0216] The inventors of this invention have provided a detailed description of the embodiments described above. However, this invention is not limited to the above embodiments and can have various variations without departing from its spirit. For example, in terms of customer terminals, in addition to smartphones, it can also include various forms of customer terminals such as laptops and tablet computers.

[0217] Furthermore, the above embodiments have provided a detailed and easily understandable description of the present invention. Each embodiment does not necessarily need to include all the contents described above. A portion of one embodiment can be replaced with another embodiment, and one embodiment can be added to another embodiment. Alternatively, adjustments such as additions, deletions, or replacements can be made to portions of each embodiment.

[0218] Furthermore, some or all of the aforementioned components, performance, and processing units can be implemented using hardware (such as integrated circuits). Additionally, the explanations and descriptions of the programs implementing processor performance by the aforementioned components, performance, and processing units can also be implemented using installation software on storage devices such as networks or disks, or network application software from application service providers. The programs implementing each performance, desktop and folder information, etc., can be stored in recording devices such as memory storage, hard drives and solid-state drives, or recording media such as IC cards, SD cards, and DVDs.

[0219] Description of Reference Numerals

[0220] 100… User Terminal

[0221] 110… server

[0222] 120… Management Server

[0223] 130…User Information Storage Department

[0224] 140… Personal Goods Database

[0225] 150…Correlation Coefficient Information Analysis Department

[0226] 160… Search Department

[0227] 170…Purchase Interface Generation Department

[0228] 180…Purchase Interface Provided by Department

[0229] 800… Search interface

[0230] 1800… Purchase page

Claims

1. A management server connected to user terminals and servers storing large amounts of data via a network: It has a correlation coefficient analysis department that analyzes the relevance coefficients of products based on the search patterns contained in the search history provided by the user information storage department, and extracts the relevant product information of the products involved in the analysis from the server. The correlation coefficient information analysis unit updates the information by covering the correlation coefficient information contained in the extracted product information with the correlation coefficients of the analyzed products. The aforementioned analysis unit also has a personal product database that has been stored before updating product information and a search unit that extracts relevant product information corresponding to the acquired user information from the aforementioned personal product database; The search department generates a search interface based on the attribute data and images contained in the extracted product information. The search unit will determine the target benchmark product based on the user's input of desired products received from the search interface displayed on the user terminal; in addition, the search unit will also determine additional conditions based on the user's input of filter tags received from the search interface displayed on the user terminal; and extract similar product information from the personal product database based on the target benchmark product and the additional conditions. The search department generates a search interface based on the extracted product information. The aforementioned search unit will provide the generated search interface to the aforementioned user terminal; The similarity is determined based on whether the parameters of each item in the product information attribute data are the same or similar, or how many parameters are the same or similar, and it increases with repeated searches. The relevance mentioned above is determined based on the relevance coefficient of the product information, and it increases with repeated searches. The correlation coefficient mentioned above is a value ranging from 0 to 1, which determines the relative correlation between each product and other products.

2. According to the management server of claim 1, the search unit allows the user to determine whether to change the target benchmark product or additional conditions; if not, the search unit extracts product information from the personal product database that maintains a similarity level 2 to the target benchmark product and additional conditions, and is other than the content already provided to the user terminal; the search unit generates a search interface based on the extracted product information. The search unit provides the generated search interface to the user terminal; then the user decides whether to perform a further search; if a further search is performed, a search interface with similarity of n+1 is generated in the nth further search.

3. According to the management server of claim 1, the search unit receives multiple target benchmark products that have been determined from the search interface displayed on the user terminal; in addition, the search unit also receives additional conditions determined by the user inputting filter tags in the search interface displayed on the user terminal; and extracts product information with a similarity of 2 corresponding to the multiple target benchmark products and additional conditions from the personal product database. The search department generates a search interface based on the extracted product information. The aforementioned search unit will provide the generated search interface to the user terminal.

4. According to claim 1, the search unit will extract from the personal product database product information that is highly relevant to the matching benchmark products determined in the search interface displayed on the user terminal and the additional conditions determined by the user input filter tags received from the user in the search interface. The search department generates a search interface based on the extracted product information. The aforementioned search unit will provide the generated search interface to the user terminal.

5. The management server according to claim 1, wherein the search unit determines the single benchmark product by having the user drag and drop the package product icon to the center grid in the search interface displayed on the user terminal; then the search unit extracts product information with high relevance corresponding to the single benchmark product from the personal product database; The search department generates a search interface based on the extracted product information. The aforementioned search unit will provide the generated search interface to the user terminal.

6. According to claim 1, the search unit will not extract product information from the search interface displayed on the user terminal from the personal product database; The search unit deletes the content displayed in the center cell of the 3x3 grid and generates a search interface that randomly displays products in the cells outside the center cell; The aforementioned search unit will provide the generated search interface to the user terminal.

7. The management server according to claim 1, wherein the management server receives a desired product determined by the user clicking on a magnified display tag included in the search interface displayed on the user terminal; extracts the product information of the desired product from the server; generates a purchase interface based on the extracted product information; and provides the generated purchase interface to the user terminal.

8. The management server according to claim 1, wherein when a user receives an instruction to search again or purchase a product by clicking on a re-search tag or zoom-in display tag included in the search interface displayed on the user terminal, the management server has a function for charging for banner advertisements.

9. The product search method for a management server as described in claim 1, wherein the product search method comprises the following procedures: The program that the management server obtains and stores user information from the user terminal includes the following steps: First, the management server analyzes the relevance coefficients of products based on search patterns contained in the search history provided by the user information storage unit, and extracts product information related to the analyzed products from the server. Second, the management server updates the stored information by overwriting the extracted relevance coefficients with the analyzed product relevance coefficients. Third, the management server extracts product information from the personal product database that corresponds to the obtained user information. Fourth, the management server generates a search interface based on the attribute data and images contained in the extracted product information. Fifth, the management server extracts product information from the personal product database that corresponds to the target benchmark product determined by allowing the user to select the desired product in the search interface displayed on the user terminal and the additional conditions determined by the user's input filter tags received from the user in the search interface. Sixth, the management server generates a search interface based on the product information. Finally, the management server provides the generated search interface to the user terminal.

10. The product search method for the management server according to claim 9, wherein the product search method comprises: The aforementioned management server will ask the user to determine whether to change the target benchmark product or additional conditions; if not, it will extract product information from the personal product database that maintains a similarity of 2 to the target benchmark product and additional conditions, and is not included in the content already provided to the user's terminal. The aforementioned management server generates a search interface program based on the extracted product information. The management server then provides the generated search interface to the user terminal program, which asks the user to determine whether to perform a further search. If a further search is performed, a program generates a search interface with a similarity of n+1 in the nth further search.

11. The product search method for the management server according to claim 9, wherein the product search method comprises: The aforementioned management server is a program that extracts product information with a similarity score of 2 from the personal product database, which corresponds to multiple target benchmark products determined in the search interface displayed on the user terminal and additional conditions determined by the user's input filter tags received from the user in the search interface. The aforementioned management server generates a search interface based on the extracted product information and provides the generated search interface to the user terminal.

12. The product search method for the management server according to claim 9, wherein the product search method comprises: The aforementioned management server is a program that extracts highly relevant product information from the personal product database, corresponding to the benchmark products determined in the search interface displayed on the user terminal and the additional conditions determined by the user's input filter tags received from the user in the search interface. The aforementioned management server generates the search interface based on the extracted product information and provides the generated search interface to the user terminal.

13. The product search method for the management server according to claim 9, wherein the product search method comprises: The aforementioned management server extracts product information from the personal product database that is highly relevant to the individual benchmark product in the search interface displayed on the user terminal by allowing the user to drag the set product icon to the center grid. The aforementioned management server generates the search interface based on the extracted product information and provides the generated search interface to the user terminal.

14. The product search method for the management server according to claim 9, wherein the product search method comprises: The aforementioned management server does not extract product information from the personal product database from the search interface displayed on the user terminal. Instead, it deletes the displayed content in the center cell of the 3x3 grid and generates a search interface that randomly displays products in the cells outside the center cell. Finally, it provides the generated search interface to the user terminal.

15. The product search method for the management server according to claim 9, wherein the product search method comprises: The aforementioned management server receives a program indicating the desired purchase item determined by the user clicking on the magnified display tab in the search interface displayed on the user terminal. The aforementioned management server extracts the product information of the desired purchase item from the server. The aforementioned management server generates a purchase interface based on the extracted product information. The aforementioned management server provides the generated purchase interface to the aforementioned user terminal.

Images